The more I live, the older I get, and the more the world changes around us, the more I see that our minds are not designed well for the current complexity of life in modern civilizations.
Our minds are built for very slow-moving and almost static environments where the people you know, the tools you use, and the dynamics around you change at a very glacial pace. The pace used to be so slow, that it would take generations for things to change. A ancient person could move back and forward in time 500 years and they'd likely be in a world that's mostly the same - same farming tools, same modes of transport, same deeply-local and family-based existence.
Imagine you went forward 500 in time. You're certain -- as I am -- that whatever world we arrive at 500 years from now, it'll be vastly different from the world today. We'd feel genuine anxiety traveling 500 years into the future because of how unpredictable we know it could be. We know at a deep subconscious level that the world is changing quickly.
This cosmopolitan, global, hyper-dynamic, tech-filled, and complex world we live in today requires that we embrace technology and use it in productive and effective ways. By using tech to offload memory, for example, we'll be making a prudent decision - remembering a few things in ancient times is different than having to remember hundreds of things with each thing in its own category (eg. home, work, car, finances, etc.). Another example might be using finance software or MS Excel to manage your money - living in a world with no money at all and mostly living in a quasi-communist family-oriented society is quite different than having to manage multiple account and making transactions with other humans and organizations on a daily basis.
It's fun and easy to focus on macro stuff - we read or watch the news and see what's happening in DC and around the world. The media loves to write about such things because it gets readers/clicks and it's something everyone is going to find relevant (as opposed to micro stuff like your local real estate market, which is not important to 99.99% of the world).
In reality, however, the micro stuff is more important most of the time. Not all the time, of course - sometimes macro events can have such large (usually negative) impacts that they outweigh anything that is happening at the micro level. In these cases, you hopefully have built a strong and resilient existence to weather any storms that may have come your way.
The micro things are your day-to-day habits and your ways of living in this world - they may include things like
One would be well-served by focusing on the micro stuff while ensuring they are well-protected from significantly adverse macro events.
How to use historical stock market data to build your investing intuition, and move one step closer to becoming an investing superhero
On Monday, October 17, 1987, the S&P 500 dropped a little over 20%. Going back to 1950, this was the single worst one-day drop in the stock market. It showed and taught a lot then, and it can still teach market participants and investors today if they are willing to listen.
Historical financial data can be magical - it can help you travel into the past and see the forests from the trees. By listening to historical data, we can more easily understand that a single-day drop of 20% in an index such as the S&P 500 is possible. By knowing that a 20% stock market drop is possible -- and by seeing the number present itself to us in the data -- we can better understand the risks we face by investing and participating in the financial world.
How do we know whats or of stock market drops are possible?
What exactly does possible mean in the context of investing and considering severe market declines? Sure, we know it's "possible" for the S&P 500 to drop by this amount or that amount. But until you ground your understanding in some historical data, you're not going understand this possibility at a deep level.
Observing the S&P 500's daily price movements will help you learn what sort of severe drops are possible for your portfolio
First, a quick note on using the S&P 500: The S&P 500 is a great place to start because it's a reasonable proxy for the market. The S&P 500 surely doesn't represent the entire market of equities, financial assets, and let alone of all assets; but, it's a reasonable and easily-manipulatable proxy for "the market."
If we care about how bad things can get in a single day (eg. an extremely severe yet plausible one day decline in your portfolio), we'll want to look at daily price data. This type of data is readily available online. There's a lot of free data, but you'll have to pay to access longer time horizons or more esoteric data.
Below, we have S&P 500 data April 3, 1950 to September 6, 2019. This represents an almost 70-year time horizon, a bit less than the expected lifespan of a person today. You can see this in the screenshots below (bottom and top of the table shown; table ordered earliest to latest).
The data has three columns - one shows the date, the other the closing price of the S&P 500 on that date, and the last column is the day-over-day change int he S&P 500 stated as a percent. The day-over-day change is easy to calculate - it's merely the one day change divided by the previous day's closing price. Declared as a formula, it is: [Day 2 - Day 1]/Day 1
This is called the arithmetic return. More complex return types -- namely the geometric return -- exist, but they are outside the scope of this discussion. The simple arithmetic return above is sufficient for our purposes.
Observing financial market data can teach you a lot and help to build a bit of market-related intuition
Just observing stock market data like this can be useful. Exploring data visually without graphing it can give us some interesting and potentially-valuable preliminary insights. This is especially true for people who haven't done this sort of data analysis before. For example, we can observe that in the very beginning of our data set (the first few days of April 1950), the S&P 500 was around $18. This is in sharp contrast to the almost $3000 S&P 500 level we observe below for September 2019. This plainly shows us that there has been some dramatic growth over the last 70 years in absolute metrics.
Although we can see some things by observing the financial data, it's hard to determine summary statistics about the S&P 500 data set visually. Stats like mean, median, minimum, and maximum are hard to see because all of the data needs to be taken into account. Taking all of the data into account can't easily be done relying on the human mind - it's just not what it's made to do. The data set has almost 17,500 rows - that's simply too much to comprehend without the use of computing devices/methods (or without a considerable amount of time to devote to this endeavor). Luckily, such devices/methods are easily available for free (eg. Google Sheets) or cheap (eg. Microsoft Excel) in the form of software. More complex options are available that are both free and paid (eg. R, Matlab, Tableau, etc.). Something like Microsoft Excel would be enough for the vast majority of use cases, however.
Finding the minimum, or the worst stock market day over the last 70 years
Some relatively easy functions can be used in Excel -- the tool of choice for most finance people -- to get some stats on the data. In the screenshot below, you can see the average, the max, and the min. The min is what we care about most here - it represents the lowest day-over-day S&P 500 change; it represents the most severe single-day drop in the S&P 500 over the last 70 years.
We can see that the worst drop is -20.47%. That means that in one single day, the stock market effectively dropped by over 20%.
Black Monday - an over 20% one day drop in the stock market
We can see the drop occurred on Monday, October 19, 1987, by examing our data set in greater detail. By filtering the data from largest to smallest, we can see what date corresponds to the worst stock market drop. Once we have the date, we can observe what happened around it in the days before and after Black Monday, which is what Monday, October 19, 1987is called in the financial industry.
The image below is the copied and pasted data from around Black Monday. It's interesting and useful to observe what happened around that time. We can see in the 10 days around Black Monday, 7 out of 10 days were losing days. We can also combine the losses to see what the cumulative loss over the 10-day period would have been. We can see that it's even worse - over the 10 days, the stock market dropped over 26%.
That means you could wake up one day over the course of your investing life and see that your portfolio is down 20% in a single day. That event would be tough to deal with - you'd be in for a very rough day and rough week. It would take some time to recover from the loss, but recovery would definitely be possible. A mistake, however, would be to panic and deviate form your long term investment strategy for no real reason beyond the fact that you're freaked out.
Use this knowledge to avoid panic sales and other forms of freaking out during the next inevitable stock market disaster
We all get freaked out in investing - it's your money that's on the line, and you don't want to lose it. Even small drops can seem bad. Even times where there's no movement could be perceived as bad if you were anticipating gains. You can't let these investing difficulties make you make investing mistakes, however. You've got to do your best to maintain a long-term perspective on investing. Something that helps us do that is exercises like the one we just went through. Looking at historical market data, understanding how the markets have moved over time, and understanding how markets may tend to move int he future are all essential things that will buffer you from foolish investing mistakes made out of fear.
If you'd like to explore the Excel file form which the above screenshots originated, you can download the file here. You'll be able to copy and paste the S&P 500 data to do your own data analysis work, like finding the maximum one-day increase.
Can investing in your career provide more financial benefits than investing in the financial markets? Maybe...
For most people -- especially for those in the 1st half of their working lives/careers -- investing time, energy, and even money into their careers might prove to be very valuable investments. For many, investments in their careers might prove more rewarding than typical financial investments.
The reasons why this is the case for many people are two-fold. First, most people don't have a lot of initial capital to start off with - saving and investing is key for them, but with so little capital there's not much that can be achieved in the short run for the typical retail investor. Second, investments in your career (e.g., investments in your skills and knowledge) have compounding effects over time as one progresses in his/her career.
It's hard to give a useful guide on what to pour your money, your time, and your energy into because each job is different and careers are diverse. However, it's a safe bet that building the following skills/attributes within yourself will prove very beneficial over time:
Newton said there's inertia in the universe, so we now know more about our world and physics is better for it. That's not the inertia we're talking about here. Forget the universe for a second - focus on inertia in your mind.
Mental or spiritual inertia is a real thing. We won't try to define it here, but everyone who has experienced it knows what it is. It's when
Here are some examples of using interim in your own favor and taking quick, small, but intense bursts forward in whatever you'd like to achieve:
The problem with modern Western self-improvement and self-development thinking is that it treats the human mind as a machine when it should instead be treated more like a tree - we'll get into what this actually means in below. But, the vital thing to note is that this type of thinking has permeated self-improvement and self-development thinking quite profoundly. It has penetrated so deeply that when most people think of becoming better human beings are strictly in the machine paradigm; most people don't even understand that a different way of thinking about the mind and self-improvement exists.
Machine vs. Tree: Be a tree, not a machine
Machine: The machine paradigm is easy for most Western readers (e.g., readers who grew up in an environment where Western post-Platonic thought formed the foundation of academic/scientific thought) to understand. Treating your mind like machine means having a paradigm where you believe improvements to the machine (your mind) are to be made based on external analysis/planning/thinking (exogenous improvements) and where those improvements can be immediately implemented (e.g., upgrading the machine).
Tree: The tree paradigm is more difficult for Western-oriented thinkers to understand and is somewhat more in line with Eastern, though, but not completely. The tree paradigm is where you believe that mental "upgrades" are impossible or exceedingly rare, and you acknowledge how little control you actually have over your own mind. Instead, with the tree paradigm, you more clearly see the only real way to make lasting changes to your mind: through feeding it with useful information over a long period of time and allowing that information to be absorbed, integrate, and recalled later. Someone who understands the tree paradigm has a far clearer perspective on their own mind, their ability to improve or develop it, and the timeframes it takes for such improvements. This is just a brief into the tree paradigm - there's as much and more here as there is in the well-known machine paradigm
The descriptions above are accurate, but they might seem confusing to readers without proper examples. Often, the best way to illustrate a point quickly is to give examples. So, here are a few.
Problem: A man realizes he's terrible at relationships - his wife is unhappy and he finally realizes that there are things he just doesn't know about women, relationships, and how to have a happy marriage.
Problem: A man's friends sit him down and tell him that they feel he has a deep problem with aggression - at bars he picks fights, friends are always afraid of him getting upset when he's drunk, and they remind him of how he became aggressive with his wife a few months ago.
Problem: A high school kid who is good in school but self-conscious, timid, and possibly under-developed physically compared to his peers gets harassed at school by older, more aggressive kids looking for easy prey.
News is by definition most relevant in the short term – as time moves forward each piece of news information degrades quickly in terms of how relevant and/or useful it is. In effect, information can be thought of as having a half-life. If we map things out in this respect, we can see that not all information is equal:
So, if you’ve got a limited amount of time and energy -- and your goal is to maximize the amount of useful information you obtain -- you’ll want to focus on things with a far lower half-life. In practice, that means making choices like this:
Here are two great articles that in-part inspired this piece:
Once you understand the what the beta of a stock is and what it represents from a finance and a risk perspective, the next step is calculating it. Reading the theory behind the beta is important, but at some point actually calculating a stock's beta for yourself will prove more useful than reading another paragraph of finance theory. Here, we'll walk through a basic example of how to calculate the beta using MS Excel. We'll use the S&P 500 as a proxy for the market and we'll calculate the beta of Facebook (FB) stock.
Step 1: Obtain Daily Stock and S&P 500 Prices for 1 Year
The first step is to obtain daily prices of both the S&P 500 and the stock we're looking at (Facebook in this case) for a period of 1 year. We'll want at least one year's worth of prices in order to capture a full year's economic cycle, include all of the seasons, holidays, and any unique things that might influence the market over the course of a year in our data set.
We'll want to make sure that the dates align - we want to make sure that for every day we have both an S&P 500 piece and an FB price. Basically, what you want to avoid is a situation where you have the S&P 500 price for a day but don't have the FB price for that same day (or vice versa). This should be easy if you're using stocks from the US as bank holidays will generally coincide.
A final point to note is that you'll want the adjusted closing price for each day (as opposed to the general closing price). The adjusted closing price takes things like stock splits into account. Imagine a stock split occurred for the stock you're analyzing halfway through your yearly timeframe - this would make it seem like there was a huge price drop. To avoid this, adjusted closing prices (which are readily available online alongside normal historical closing prices) take this into account and provide (usually) a post-split price for the entire timeframe.
As to where to obtain the data, that should be easy in today's world - you can go to any of the major finance websites to download historical data or you might use your own brokerage account's platform. You can also use a paid data provider, but that's not a necessary expense for most people.
Step 2: Get the Stock and S&P 500 Price Data Neatly into One Excel File
Next, you'll want to copy the data into the same Excel file so you can work with it. This should be easy. Take care to leave a few columns between the data so that you can do the next calculations we'll go over below.
Step 3: Calculate the Daily Returns for the Stock and the S&P 500
Next, you'll want to calculate daily returns (eg. daily price changes) for the S&P 500 and FB. This can be done in two ways:
simple return: (today - yesterday)/yesterday
log return: ln(today/yesterday)
For the most part, simple returns will suffice. Sometimes log returns are useful because they inherently assist with normalizing the data, but that's both beyond the scope of this discussion and unnecessary for us here.
It's easy to calculate simple daily returns in Excel (see the formula in the image below). Once you have a single cell filled out, you can drag the cell all the way to the bottom to create a time series of daily returns for both the S&P 500 and FB.
Note that for the very final day (here it will be the first day in our time series), you won't be able to calculate a return - you'll get an error message in Excel. This is because you won't have a price for the previous day and you'll effectively be dividing by zero. This is not relevant for our purposes and this can safely be ignored.
Recall that the beta can be calculated by using the following formula:
beta = cov(x,y)/var(x)
where x is the stock and y is the S&P 500 and where var(x) does not equal 0.
So, we must calculate two things:
You can see this in the below images - notice the highlighted formulas and the sections of the Excel sheet they reference.
Finally, you simply divide the two obtained numbers per the above formula - notice this in the image below where we divide the obtained covariance by the obtained variance.
We now have our beta for FB - it's 0.861 as of the end of February 2017 - remember that this can change as the market changes and as Facebook changes. We'll notice that the beta is less than 1 - this means that Facebook stock is less volatile than the market (as represented by the S&P 500).
Step 4: Calculate the Two Subcomponents of the Beta Formula
A Bit of Intuition Building - Let's Graph the Stock Movements
We now have our beta, but let's go even deeper to build some intuition around the number. Below, we have created two portoflios, each consisting of $10,000 - at the outset, we invest the full $10,000 in either the S&P 500 or Facebook. So, at the beginning of our time series (February 26, 2016), we have the following:
We're doing this because we need to somehow compare the prices - if we only look at the movement of one share of the S&P 500 vs one share of FB, we won't get a clear picture because the starting numbers are different. What we care about is not the absolute amounts, but the relative movements of both.
Below you'll see a graph of how the portfolio would have moved throughout the year - this is literally what would have happened had $10,000 been invested as we described above. Here we see some interesting things:
An introduction to Beta, a foundational concept in finance that measures exposure to general market risk (systematic risk)
A precursor to truly understanding the concept of Beta is an understanding of the difference between systematic (unverifiable) and unsystematic (diversifiable) risk. These terms sound complicated, but they really aren't.
Here's a very brief review of the difference between these two types of risk:
Keeping in mind the above definitions, it would be useful to know how much systematic risk you are being exposed to with a given security (eg. a stock) relative to the market as a whole. Stated another way (and hopefully more simply), if you're going to hold a stock, it's useful to know how risky that stock is relative to the market - this knowledge will allow you to understand how the stock will fit into an already diversified portfolio and it will also allow you to use the important Capital Asset Pricing Model (CAPM) down the line. Stated yet another way for the purposes of clarifying a possibly confusing topic, the beta of a stock will allow you to know the market risk of the stock (the risk arising from general market factors) - it is not, however, a measure of the idiosyncratic risk fo the stock.
Boats, Water, and Rain: An Example to Help Clarify the Meaning of Stock's Beta and Why We Care About It
A question that arose for me in studying finance was in this effect:
If we are only looking at systematic (eg. market risk), why would any stock have a different exposure to market risk? If finance theory says that there is a certain risk called diversifiable market risk that you're still exposed to even if you have a diversified portfolio, why would betas be different? Shouldn't all betas be the same?
Stated another way, this question might sound like:
Why does beta tell you about systematic (market) risk and not unsystematic (idiosyncratic) risk - and why does that even really matter?
This question evinced a deep lack of understanding of finance and further study clarified things for me enough so that the question itself seemed foolish. I will attempt to provide context here so that such foolish questions don't arise.
Let's leave finance altogether and travel to a port. In that port, there are wooden boats of all shapes, sizes, and designs on the shore. The port is open and you can take any one of them and go out into the water. The port has constant clouds overhead and there is constant rain. This is a unique port b/c the rain is different in different places - if you're standing on the beach and you walk just 10 feet to another direction, the amount and strength of the rain will change.
Now, imagine you take a boat out into the water. You'll feel the rocking and the swaying of the water. No matter where you are in the port area, you're going to feel the water. If a heavy wind storm comes in, all boats will be affected. If it's calm today, all boats will be calm. The way your boat feels, however, is going to be based on the design of your boat - a large heavy boat will sway less than a small boat and a swiftly designed boat that can cut through the water will react differently than a rugged boat. Before you go onto the water, in anticipation of the chaotic rain in this unusual port, you can easily construct a quasi-roof over your boat to totally protect you from the rain. You can choose to go out with no roof at all and be totally exposed to the rain. You can choose to go out with a poorly made roof and just have limited protection against the rain. You can also choose to go out with a fully-built roof and be totally protected from the rain.
You go out into the water now and some wind comes in. No matter what you do, your boat will be affected by the water moving. This is like systematic risk - all boats are affected by it. This risk is not diversifiable - no matter what you do, if you're in the water (eg. if you're in the market), you are exposed to this general risk of moving waters (eg. general market risk). So, if a systematic risk is moving water, a unysystematic risk is the rain - you can choose to diversify that risk away by simply putting up the roof we discussed earlier. You don't have to be exposed to it and many people on the water probably aren't because they've put up roofs - this risk is idiosyncratic to each boat and is diversifiable.
In choosing a boat, therefore, you might want to think about a few things. For one, you might want to decide if you want a roof up. Another important thing to think about is the shape of the boat. Since you know the water will move and you will always be exposed to that movement, you'll want to know how your boat will move relative tot he movement of the overall water. You'll want to look at each boat and think about whether or not it will move calmly or forcefully whenever the water moves. Forceful movement isn't necessarily bad, but you'll still want to know what kind of a journey you're about to have.
This looking at the boat and seeing how it will react to movements of the water is exactly what the beta is about - knowing a stock's beta will allow you to know how the price of that stock will move relative to moves in the market overall. Clearly, the beta then will not tell you about idiosyncratic risks (just like the shape of the boat will not tell you whether or not you'll be exposed to the rain - that's something for you to decide based on your diversification). It's not up to you to control as stock's beta just like it's not up to you to control how the boat will move - the boats are there laid out for you and built already, you can simply choose a pre-built one.
Going further, we now can see that just because the water moves a certain way (eg. just because the market goes a certain way), it does not at all imply that each stock will move the same way. It is obvious when we think of our port - no one would ever question that boats designed differently would move differently in the water. By the same token, it should be easy to see that firms (which are comprised of different people, processes, assets, liabilities, products, knowledge, etc.) will move differently when the overall market shifts and changes.
Why do firms move differently with the market?
Going a bit further, we can ask what the underlying causes are for different betas (for different movements relative to market risk). We know why the boats move different (because they are designed differently), but understanding why firms move differently is a far more complicated matter.
Firms are different in many ways. Some of these ways include:
These things, and much more, clearly influenced the way a firm's cash flows and stock price (which is dependent on both cash flows and overall market sentiment int the short run) will change based on changes in the market. A firm located in a single US state in Middle America that sells basic goods to people of that state exclusively is clearly exposed to less market risk (eg. less geopolitical risk, less market risk, exchange rate risk, economic downturns, etc) than a multinational firm that produces services that businesses generally purchase in prosperous times but can do without in difficult times. Clearly, one firm's beta would be less than the other if the beta is a measure of a firm's relative exposure to market risk.
Quick Note on the Beta of the Market
The beta of the market (eg. the beta of the S&P 500) is said to be 1. This will make sense further down the line because we will see that the beta is calculated by seeing how a stock's prices more relative to the S&P 500. A beta greater than 1 indicates a stock more volatile than the S&P 500 and vice versa - clearly, then the S&P 500 will have a beta of 1 because it moves with perfect correlation to itself.
Another Quick Note on What Diversification Means
Now that we've gone through the basics, we must define diversification. True diversification when discussing beta and when saying that an investment is diversified involves holding the entire market - meaning all stocks, all bonds, all real estate, etc. (or a portion of a basket of them). Holding just a diverse portfolio of stocks still exposes the holder to idiosyncratic risk - they are exposed to shocks that only or primarily affect the stock market.
Now, it is difficult and cumbersome to use as total market basket - one doesn't really exist because it's hard to value things that don't have regular market prices like stocks. Therefore, most finance texts use a proxy for the market - that proxy is the S&P 500 most of the time. We'll note that this is a weak proxy because it only focuses on 500 major US stocks - ignoring all of the other stocks and asset classes that one could invest in. Keeping that in mind, we can proceed with using the S&P 500 due to the fact that it is commonly used and that it will still produce reliable results and metrics for our use and understanding.
And Finally...How to Actually Calculate the Beta
We've spent a lot of words and sentences discussing what the beta is, but without an actual walkthrough of the calculation, the entire concept is likely to still be obscure to those who have not studied finance before. Let's dive right into the calculation.
Another way to take the beta is the correlation of price movements of a stock to price movements of the S&P 500 (eg. the market). We talked about risk before this, but in order to actually quantify these concepts, we must move from a world of words to a world of numbers. We can do that by talking about prices - risk can be represented by volatility (by price movements). We can look at the price movements of a stock and the price movements of the S&P 500 side by side and see how the move - they might move in tandem, the stock might move more aggressively than the S&P 500 (more volatile - higher beta), or it might move more calmly than the S&P 500 (lower volatility - lower beta).
In order to get the numbers (both the stock you're looking at and the S&P 500), you can simply use the internet to obtain historical prices - it's a relatively simple exercise. You'll want daily prices and you'll want to make sure that the data lines up in terms of day and the exclusion of weekends - you'll want a day-to-day match up. Additionally, if any stock splits occurred over the period you're looking at, you'll want to take the post-split stock price for the entire time period - this will avoid adding a lot of error because if you don't adjust the price it will seem like the price dropped significantly on the day of the split. It's pretty easy to use a post-split number for the entire period (post and pre-split) because most online repositories of historical stock data will do this for you automatically.
Next, you'll have to calculate daily returns - this is simply done with the follwoing formula and should intuitively makes sense:
return = (today's price - yesterday's price)/(yesterday's price)
This is a simple percentage change over a single day and this should be done for both your stock and for the S&P 500. You'll now have a list of daily price changes for both the stock and the S&P 500. The reason we look at price changes is because we want to see how the movements of the stock related to movements of the s&P 500 - in effect, we don't really care about the absolute prices of either the stock or the S&P 500 but are only concerned with their movements over time.
One important thing to note is that you'll want to capture enough time within your analysis - you'll likely want a full year's worth of data.
Finally, you'll compare the price movements of the stock to the S&P 500 using the most commonly used formula for calculating the beta of a stock:
beta = cov(x,y)/var(x)
where x is the stock and y is the S&P 500 and where var(x) does not equal 0.
This formula might seem complicated, but it really isn't - it can be easily explained and easily performed din a program such as MS Excel using the functions COVAR() and VAR() over a list of price changes.
Covariance (which is represented by cov in the above formual) is simply a measure of the joint variability of two random variables - it's a measure of the degree two random variables (here the random variables are the price changes) move in tandem with each other. Variance (which is represented by var in the above formula) is simply a measure of how much a random variable (here the random variable in question is S&P 500 price changes) moves about its mean.
Once you have the covariance of the stock and the S&P 500 and the variance of the S&P 500, you simply divide the two numbers per the above formula to obtain the stock's beta - you now have a very powerful piece of information telling you how a stock moves relative to the S&P 500 (which is a proxy for the market). You now know how risk the stock is relative to the market and how much risk the stock would add to a diversified portfolio - you now know the systematic (undiversifiable) risk of the stock.
The 5 types of beta
We touched on this above, but let's formally review the possible betas:
There are No "Bad" Betas
Remember, in no place did we say that any beta measures are bad. A beta of less than 1 is not bad. A beta of greater than 1 is not bad. Beta simply tells you how much the price of a stock varies relative to the market, it does not imply anything beyond that. A beta of less than one might be desirable for a conservative investor while a high beta might be desirable for a more aggressive investor who is looking for more return. A stock's beta tells us a bit about the expected turn of the stock with higher betas indicating higher expected returns - this makes sense because more risk should entail more reward. However, this discussion about translating beta measures into an understanding of a stock's expected return is beyond of the scope of this present discussion.
The Complete History of the Dow: The changing companies that made up the Dow Jones Industrial Average since the prominent stock index's inception
The Dow Jones Industrial Average is one of the longest-running stock market indexes in the world. Its components have changed since inception - they've changed 51 times since the inception of the index by Charles Dow.
Looking at the Dow Jones Industrial Average's (or simply the Dow's) components over time allows us to see how American business (and the world in general) has changed over the last century and a half.
We won't go into all 51 component changes here -- you can find them elsewhere if you'd like -- but we will focus on the most interest and relevant ones and discuss them in a bit more depth than you can find elsewhere on the internet. Instead of giving a cursory overview, we'll dig a bit deeper to see what underlying changes were the root causes of the changes and in the process, we'll gain the following benefits:
The Dow on July 3, 1884 (precursor)
The initial Dow (which wasn't properly the Dow Jones Industrial Average but was instead a creation of Dow called the Dow Transportation Average) consisted of the following:
As the original "Transportation Average" name should indicate, the original Dow components were heavily focused on transportation. We can clearly see that there are a lot of railroad companies represented in the initial Dow mix. In the 1880s, railroads had been around for a few decades, but they still represented the new and happening industry - similar to how technology is today fast growing and focused on thing in business even though computers have been around for a few decades already. Railroads represented Manifest Destiny and a new industrial era where lots of money was being made in the business of moving things from one place to another.
We see that 9 out of the initial 11 firms represented in the 1884 Dow were railroad companies - that's a very large representation and should clearly indicate the importance of transportation generally (and railroads specifically) in the pre-20th Century US economy. As the country moved westward and as more and more goods were in need of rapid transportation in the post-Industrial Revolution era, railroads were able to extract very healthy nominal and real profits.
Basically, a discussion of the early years of the Dow inherently is a discussion of railroads. The first public railways opened up in the US in 1830 using steam engine - by the 1880s, technology had improved as did ridership and a need for transporting goods in a new type of economy where self-reliance was beginning to give way to mass consumption and production.
The equivalent today in terms of industry would be seeing all tech firms dominating the Dow Jones Industrial Average - imagine seeing the Dow today composed of the likes of Google, Facebook, Microsoft, Oracle, Salesforce, Twitter, Apple, HP, Dell, Cisco, etc. An observer would think that the US economy was heavily dominated by tech. Luckily for us, today's economy is far more diverse than the industrial and transportation economy of the late 19th century - we have large industrial firms, firms involved in chemicals, firms involved in telecommunications, firms producing basic products, firms that primarily provide services (eg. consulting firms), etc. Today's economy is as diverse as any has been in human history.
May 26, 1896 (the first proper Dow Jones Industrial Average)
The first proper (non-transportation only) Dow Jones Industrial included the following firms:
This was the first real Dow Jones Industrial Average. Here we see many of the railroad companies replaced - only two of the firms (the Northern American Company and the Tennessee Coal, Iron and Railroad Company) are firms heavily involved in transportation.
We can see that the list has now moved away from transportation and is focused important necessities for late 19th Century America. Things like cotton, oil, tobacco, cattle feed, coal, iron, leather and rubber are all represented - these basic necessities were key to a life that was moving away from self-reliance on farms and into a mass-produce economy that required energy (in the form of gas, oil, and coal), straps, linens and other fabrics, heavy metal, electricity, etc.
If the Dow had existed 500 years prior in the Middle Ages, things like electricity, leather, cotton, coal, and rubber would not be there - most of life would consist of cattle feed and other types of feed.
Another interesting thing to note is that the names of these firms are quite basic - they are literally are names of what the company produces. Can there be any doubt that the Tennessee Coal, Iron, and Railroad Company is involved in the production of coal, iron, and railroads? Would you be surprised to find out that the United States Rubber Company produces rubber? These firms were the first of their kind - they are representations of commerce and big business in an era that had only recently exited the darkness of the Middle Ages via the Renaissance. The unique names we see that not only don't represent the firm's products or services but sometimes are not even traditional words that humans have used are only possible in a world that understands what firms are - world filled with people used to branding, buying things from companies instead of from friends or family, and have a lot of trust in business and capitalism in general. The ability of firms to market and brand themselves in order to educate the public about their products and services allows firms today to eschew the basic naming conventions of the past and to use innovative and obscure names such as Twitter or Exxon. A Twitter or an Exxon would be strange in the early years of the Dow - no one would have any idea what these firms produced. Without the ability to create an image of the firm through the use of advertising (which requires a lot - print ads, TV, radio, the internet, etc.), firms would who used strange names would find themselves at a deep disadvantage in the past. It was a far smarter idea to make sure people knew what your business did just by reading the name.
October 1, 1928 (Dow expanded to 30 firms)
As the index expanded to contain 30 firms (the size it's been ever since), the Dow was comprised of the following firms:
Here, the index was expanded to the 30 firms we have today. This was an interesting time in the history of the United States and especially its economic history. The Roaring Twenties were coming to a close and little did anyone knows that the Great Depression was right around the corner.
Here we can see the that we have a few automotive firms represented - we've got General Motors, American Car, Mack Trucks, and Nash Motors. Car companies have come on the market and are now some of the largest firms in the country. A car firm at this time would be similar to seeing the edition of technology and internet firms in the 2010s and 2020s - the firms came up over a few decades and finally took their place among the largest in the US by playing in a new and important industry.
We see that the names here are still those basic names that hearken back to an era before sophisticated marketing and advertising and before readily available means of communicated such as radio, TV, and mass color print.
July 3, 1956
This is the first Dow changes after the US entered WWII - the previous Dow adjustment occurred on March 4, 1939. Since we last saw the Dow above (1928), the US had plunged into a decade-long economic downturn called the Great Depression, entered WWII (which helped it recover), and saw droves of new babies being born in post-war America (the Baby Boom). Let's see how the Dow has been affected:
Here in 1956, we can say that we are in a totally different America. The last time we checked in was in 1928 - almost 30 years later the Depression-era youths fought a war abroad and came back home to have a ton of babies. Although key staples remain in the Dow, we can see the addition of many new firms.
We can see a big variety of firms represented here - car companies, companies producing basic materials, food-related companies, retailers, energy firms, and even a photography company in the form of Kodak. In 1950s America, technology has advanced far enough to make consumer products (photography, cars, retailing, toiletries, etc.) major parts of the economy. A Procter & Gamble wouldn't exist just 50 years prior - people didn't have the disposable incomes to shower often and use toiletries nor did they have a desire to in their mostly self-reliant forms of living. In 1950s America, a firm producing household necessities would make a lot of sense. In the same light, in 1950s America, big retailers, big tobacco, and big car companies all make sense - our conception of that era is of one that has now moved way past the agrarian roots of the United States and now is in the realm of post-WWII technology and sophistication. If you had told the people living through the Great Depression that a photography firm (Kodak) would be one of the biggest in the country, they would have scoffed and not understood why - photography was a luxury and the technology was not all there yet. The same can be said about many things represented above.
August 9, 1976
Jumping forward another twenty years, let's see where this journey has brought us:
In these 20 years, surprisingly little has changed. Only a 5 firms were replaced since the last time we checked in in 1956. By comparison, there over 30 changes from 1928 to 1956 (some back and forth). What you have in this period is a stable period of growth, some merging of firms, and a movement away from those classic self-descriptive names to the more unusual firm names we know of today.
Look above to see the firms that remained in the Dow but changed their names - International Nickel became Inco, Texas Incorporated became Texaco, Swift became Esemar, and Standard Oil of NJ became Exxon. These movements are away from names that clearly state what the firm produces to more esoteric and strange names that don't provide any indication whatsoever about the firm - clearly a reliance on marketing, adverting, and branding is required in order to educate the public and create a mental picture of the firm when such strange names are used. This is possible because we are no in a world of color television, radio, print magazines, and other forms of advertising.
This trend has continued today where almost all new and interesting firms have absolutely strange names that give zero indication of what the firm actually does - names such as Twitter, Facebook, Yelp, Apple, etc. If you took a person from 1850 and asked him to whet he or she tough a firm named Standard Oil or a firm named American Can do, he or she would very likely guess correctly. If you asked the same person to describe what he or she thought a firm like Exxon does (bear in mind this is Standard Oil with a name change), they would have no clue and rightly so because Exxon is a totally made up term. As stated above, such strange names can only work in a modern world filled with modern telecommunication systems and a general populace that is receptive to advertising and marketing.
March 17, 1997
About two more decades after our last stop, a lot has happened - the Cold War is over and we're at the apex of the 1990s economic boom. Here are the Dow components now:
Here we can see more name changes (eg. Chevron and AT&T), continuing the overall movement away from the simple names to the more strange and esoteric ones that require branding.
We can also see that some of the old components (eg. AT&T, Chevron, Exxon, Union Carbide, General Electric, General Motors, Minnesota Mining, Sears, and Union Carbide) still here - times have changed but these good firms have endured for a variety of reasons. Some endured because of good management (eg. General Electric), some because of early entry and the existence of various barriers to entry (eg. General Motors), and many because of luck.
It's interesting that even though we're in the heart of the proliferation of personal computing and the internet, there are few technology firms. This makes quite a bit of sense - it takes time for these new firms to grow to a size large enough that would put them in the Dow (the top 30 firms in the US). Firms like Apple, Cisco, Microsoft, and others might have been making big moves during this era, but they were still growing. IBM and HP are present because they were around longer - IBM was around for almost a century at this time.
November 1, 1999
About two more decades after our last stop, a lot has happened - the Cold War is over and we're at the apex of the 1990s economic boom. Here are the Dow components now:
With the addition of Intel, Microsoft, and SBC, we can see that in just about 2 years, the Dow has taken on some of the tech firms. These firms have grown in market cap by now (due in part to what would later be called the Tech Bubble) and had market caps large enough to allow placement within the Dow. The Dow here contains many old stalwarts but is filled with new firms that were either started within the last few decades or came to major prominence recently.
March 19, 2015
The last Dow Jones Industrial Average change happened in early 2015 - here is the current makeup of the Dow:
In today's Dow, we see the familiar firms that make up the market share and the mind share of the US economy today. We see an almost complete transition away from the simple self-descriptive business names to esoteric and strange names that require branding - compare this final list with the first list. Firms like National Lead, Tennessee Coal, United States Leather, and others are so clear in their descriptions while firms like Visa, Pfizer, Nike, and Apple would be totally obscure if not for branding and advertising.
Another interesting thing is the rise of big pharmaceutical and healthcare firms - firms such as Merck, Pfizer, and UnitedHealth have come to major prominence due to various large-scale factors. These factors include an aging population (Baby Boomers), a more wealthy economy that can spend more on healthcare, and the success of pharmaceutical research in producing new, innovative, and expensive drugs.
We also see the tech firms playing a bigger role - Cisco, Apple, Microsoft, and Verizon are all part of the overall technology economy, helping to provide hardware, software, and telecommunication services.
Rates of return play a tremendous role in investing performance - without adequate returns, it's difficult to build real wealth
A fundamental principle of investing is that rates of return are key - but most people don't really understand their profound importance. Of course, most savers and investors know that the rate of interest they get on their savings or the rate of return they get on their investments matters a lot, but they are too easily willing to give up valuable return to things such as the below.
The common thieves of investing returns
The common thieves of peoples' investing returns have proven to typically be the following:
It's important to note that not all of the above fees are bad - you're paying these for a reason. For example, you want the mutual fund to hire a good money manager - this person will need to be compensated well. You clearly understand that administrative fees are going to exist for mutual funds and ETFs. Trading fees obviously are required so that the brokerage is paid for the service they provide you - this is a small price to pay for being able to enter and exit positions with ease.
However, you still don't want to overpay. You will not want your mutual fund or ETF to spend excessively on hiring poor-performing managers, spring money on lots of useless advertising, or running thing so inefficiently that the administrative fees are too high relative to similar funds. You'll obviously want to shop around to find a reputable and high-quality broker, but not one that charges excessive fees relative to what's available on the market. You'll also want to be disciplined and not constantly enter or exit positions so you don't accumulate excessive trading fees that will eat away at your capital. Common sense will dictate that even if the fees are reasonable in principle, they could be unreasonable in practice (meaning in amount).
To illustrate this point well, let's use an example. Examples are often an excellent way to illustrate importance finance principles in ways that are easy to understand - a theory is good but seeing numbers and graphs often allows people to really visualize the concepts being presented and gives the motivation to use the new knowledge they gained.
Investing $10k at different rates of a return - a simple example
Let's start with $10,000 in our example and let's invest that money at different rates of return - the return rates will be from 0% to 8% in intervals of 2%. First, we'll break down the possible rates and understand where you might obtain them:
Now, let's see how $10,000 will grow at each of the above rates of return by taking a look at the graph below. From looking at the graph we can see that the 0% return stays constant throughout with all of the non-zero returns separating from it more and more over time. We can also see that each 2% increase does not bring a proportional increase in the final amount - the increase itself increases over time.
The 8% portfolio brings the initial $10,000 to almost $500,000 but the 6% doesn't even reach $200,000. We can say how important even a small increase in return can make over the long term. That 2% difference is sadly something too many investors ignore. It makes sense given the human mind's propensities that a person wouldn't be able to totally and intuitively grasp the importance of even a 0.25% difference in return, but through education, we can see that the small differences end up with very big differences in results.
How can a 2% difference result in a greater than 50% difference in the final portfolio value? This doesn't seem to make too much sense at first glances - the 2% difference is only 1/4 of 8%, so shouldn't it result in a 25% difference? The maths of finance don't work this way - this is an incorrect way of thinking through it. The way it works is that the 2% you forgot on the first year doesn't stop there - that 2% you would have gained is no longer able to be around in the second year to earn additional return. For example, by forgoing 2% on the $10,000 investment, you forgo $200 in your first year, BUT it doesn't end there - in the second year that $200 would have been working for you t earn a return. The same is true in the third year, the fourth year, and so on. In effect, the person who invests at 8% is able to not only bring along that extra amount every year but to also keep that amount invested and earning. In effect, changes in investment returns compound over time. This is the underlying principal as to why small differences in return can have tremendous impacts in final portfolio value.
We aren't going deep into the maths here, but you can reference a 2013 article titled "The Arithmetic of Investment Expenses" by William F. Sharpe. The article is accessible to most readers and the title should give you a hint at the complexity of the maths - it's not very complex to calculate nad understand the impact of fees on final returns.
Next, we'll present another graph - this time with the same $10,000 initial investment but now we'll look at a broader spectrum of return rates (0% to 18%).
As we did above, let's take a look at how each of the additional returns can be achieved:
As you can see from the graph, the initial investment returns we plotted on the first graph are made to look minuscule here. Although most investors shouldn't expect to obtain returns over 14% over the long term, this graph clearly represents how important every percentage point is to the final portfolio value.
Benjamin and Gerald - How final rates of return end up mattering a lot in the long run
Finally, to really bring this home, let's go over one more example - this time let's look at two men. One is Benjamin and one is Gerald. Both Benjamin and Gerald invest $10,000 on the birth of their first child - this could be a college fund or a sort of "start of life" fund so that their progeny is financially stable. Clearly, both Benjamin and Gerald are intelligent, prudent, and caring individuals and parents - most people don't do such things. Another thing that's clear is that their children are quite lucky - they have dad's who care enough to put away some money for them at their birth. Both Benjamin and Gerald have $10,000 ready for this investment - they are quite similar in this and many respects. But, let's now see how they're different?
The strange thing is that Benjamin and Gerald are far more similar than different - in the thing that matter (caring, prudence, planning ahead, etc.), they are clearly quite similar. Their differences, as we'll see shortly, will be quite small and trivial if it wasn't for the outcome those differences would lead to.
Benjamin takes his $10,000 and invests it in a fund over the course of one year in a series of 24 purchases, once every month. He shops around for a good brokerage - the makes sure they're reputable and reliable but keeps an eye on trade pricing too. Benjamin chooses a long-term growth fund but looks at expense ratios, loads, and the quality of management in order to make sure that he's choosing the best fund for his strategy.
Gerald takes his $10,000 and invests it in 60 purchases because he is attempting to time the market. Gerald doesn't shop around for a brokerage and chooses the first one he finds. Gerald doesn't shop around for a fund, but instead takes a recommendation from his friend or family member - this fund has the same strategy as Benjamin's fund but isn't managed as well, has a load, and has a higher expense ratio.
Both Benjamin and Gerald leave the money in their account after the first year and never touch it again - they pass it down to their children who also are wise enough to leave it alone and let it grow.
Take a look at the tables above to see the actual numbers Benjamin and Gerald are dealing with. In effect, Benjamin and Gerald end up with different starting amounts and different return rates (9.75% vs. 7.25%) due to their different choices. These small differences made in the first year have tremendous impacts on the final portfolio values after 50 years. While Benjamin's portfolio is valued at over $1 million in 50 years, Gerald's is valued at only a bit above $300,000 - this is approximately a 70% difference. This 70% was a result of about a $600 difference in initial investment and a 2.5% difference in return. Most people would probably ignore these differences, but they are clearly extremely important.
If you're interested in further reading, below is a paper titled "The Arithmetic of Investment Expenses" by William F. Sharpe - a paper published by the same William Sharpe who created the famous Sharpe Ratio on how fees and expenses can impact the terminal value of a portfolio.
Average Transaction (AT) - A fundamental metric that is key to a better understanding of your business and organization
The Average Transaction (AT) is the fundamental building block to having an understanding of your business - if you don't currently know the Average Transaction (AT) for the business you own or manage, your level of business intelligence is severely lacking.
In today's works of easy record keeping, storage, and plenty of computerized analytic powers, there is no excuse to not be keenly aware of such basic and fundamental metrics such as your business's AT.
Average Transaction (AT) simply represents the mean transaction over a given period of time. Stayed more appropriately to business, AT is the expected transaction - it is the revenue you can "expect" (in the statistical sense of the term) to receive from the next individual or organization that you do business with.
Calculating the Average Transaction (AT) for your business or company
Calculating your AT is quite simple - you simply take the arithmetic average of all your transactions:
AT = (sum of transactions)/(# of transactions)
If the formula sounds very simple, it's because it is - you hopefully already know your business's AT. Of course, there are a few important things to keep in mind in order to make sure your AT is accurate and useful.
Timeframe matters when calculating Average Transaction (AT)
There are two points to be made regarding timeframe. The first is easy and hopefully obvious - you must use the same timeframe for both parts of the formula. So, if you sum the transaction over 2016, you need to divide by the number of transactions in 2016. Imagine you didn't follow this rule and instead only used 6 months worth of transaction for the top part of the formula (for the sum of the transactions). What would happen? It should be clear that you would significantly understate your AT (it works like be one-half) because you're not taking the full year's worth of transaction into account. If you only had 6 months worth of transaction, you would need to divide by the number of transactions you had in that year in order to calculate your AT properly.
The second point on the timeframe is that it's better to use a full year of data instead of just a few months. A full year of data (if your business is in a stable state) will allow the kinks and gyrations caused by changing seasons, holidays, etc. to be evened out - a full year of data will allow the full spectrum of things that occur in a year to be captured within your data.
When calculating AT, focus on transaction, NOT customers
A key part of calculation your Average Transaction (AT) is to make sure you're using transaction and not customers - it's called Average TRANSACTION after all. The distinction is key because focusing on a transaction will allow granularizing your metrics down the line - you'll be able to not just calculate AT, but you'll be able to calculate multiple ATs for different types of transactions (eg. those arising from Google, those arising from referrals, etc.). More on this is covered below, but let's look at an example to really understand the difference between using transactions instead of customers.
Imagine you have a customer that comes in once every month for a year. You'll want to count each of the 12 transactions separately instead of counting the customer as a whole. So, you'll sum each transaction and divide by 12. If you have 10 such customers, you'll sum each transaction and divide by 120 (10 x 12) because there are 120 total transactions for the year.
The benefit of doing this for transactions instead of customers can be explained by doing a thought experiment. Which would you rather have when a customer comes into your business to execute a transaction:
Clearly, the first one allows you to predict what will happen in the immediate future and put things to a close. The second one, however, only allows you to make a prediction about the overall general future - you really won't know what's going to happen today. What this example illustrates is that it's quite useful to be able to predict what's going to happen today instead of-of having today only fit into a larger long-term prediction.
Average Transaction (AT) can be broken down even further into more granular sub-metrics that will provide even greater insight into your business or firm
You can break down your AT even further to determine you AT for various types of transactions - transactions arising from Google, from Facebook, from referrals, etc.
This gradual AT is useful because it will allow you to understand where your most valuable transactions come from so that you can channel more money into those pipelines and away from less-profitable transactions.
Referrals Per Customer (RPC) - A key business metric that will allow you to learn more about your business's virality and ability to generate organic growth
Referrals Per Customer (RPC) is the correlated rate of referrals per customers over a given period of time. Stated more simply, Referrals Per Customer tells you "how much of an additional customer" each customer brings in.
Understanding that both the above definitions still might be a bit opaque and obscure to business owners and managers, let's go a bit deeper with an example. A Referrals Per Customer (RPC) rate of 0.25 means that for each unique customer over a given period of time, 0.25 (or one-quarter) of an additional customer is going to come into your business - which means that for every 4 customers, you can expect one additional customer to come in via a referral.
We measure the Referrals Per Customer (RPC) rate in terms of a single customer because it will be easier to use downstream - although it might be easier to say "you get 1 referral for every 4 customers" saying instead that "each customer brings in an additional 1/4 of a customer" is the best ay to approach and to understand RPC because it will allow you to apply an understanding of RPC to each customer and because it will be easier to use the RPC concept downstream in the calculation of things such as the Lifetime Customer Value (LCV).
To calculate your business's Referrals Per Customer (RPC) metric you simply need two numbers:
Using these two numbers, you can simply divide the number of referrals by the number of non-referrals to get your RPC metric. For example, if in 2016 you had 800 non-referral customers and 200 referrals, you would simply divide 200 by 800 to get 1/4 OR 0.25 - your RPC would be 0.25, meaning it's as if each customer brings in an additional one-quarter of a customer with him/her every time they come in.
Now that we've given you a brief overview, we'll discuss why this is an important thing to know, then we'll dive into some important conceptual pieces of Referrals Per Customer (RPC) and then follow up with an example of how to implement this new and valuable understanding.
Why should you care about Revenue Per Customer (RPC)?
Any small or medium size business owner or manager worth anything will understand the importance of referrals. From antiquity to the most modern businesses around the world today, referrals are a critical part of growing any businesses sales base - this understanding is so fundamental that it almost needs no explanation.
Humans, being social creatures, value the opinions of other humans they trust and respect. Humans intuitively understand that a referral from a respected individual is a valuable thing because it both
If referrals are so important to businesses, and if most businesses understand this, why is so little effort put into properly understanding referrals by small and medium-sized business owners and managers? In conversations with small and medium sized business owner sand managers, this usually occurs because a misconception that it is either costly or difficult to go beyond the basic "please refer us" statement to understand the nature of particular business's referrals.
If the nature of referrals can be properly understood, however, various benefits will immediately flow to the business owner or manager. These benefits include:
Correlation vs. Causation - A key distinction to know in business and in life
Now that we've covered the basics, we'll dive deeper into RPC in order to flush out some of the important details and get a good understand fo the concepts and it's potential weaknesses. First, we'll note that the way we calculate RPC is a bit flawed - RPC looks at how referrals are correlated with overall customer volume and NOT at the actual amount of referrals that a certain number of customers bring.
What this flaw means is better illustrated via a generic expamle using the same numbers we used in the brief example above. Let's say you have an ice cream shop and 800 new customer visit in 2016 with 200 referrals. Per our RPC calculation, you would look at be looking only at numbers in 2016. That means a referral could have come in on the very morning of January 1, 2016, but you would still count it as part of your RPC. This doesn't make sense because clearly, no customer in 2016 referred that customer - it was almost surely someone in 2015. So, you're not really looking at the causes of the referrals, but only at how your referrals are correlated with (eg. compare with) your non-referrals.
This is a flaw, but it should remain a minor flaw for the vast majority of businesses. You should be aware of it, but that is all - you can safely assume the flaw away because the error that will be introduced will be very small and due to the fact that the greatest error occurs in the first year. In subsequent years, although the very small error will persist within each year, the error will be normalized away via a comparison of years with each other - 2016 and 2017 could be compared with each other and both will have that error in it.
For calculating RPC, count customers, NOT Transactions
It is important when calculating your RPC metric, as stated above, to use customers and not transactions - customers might engage in multiple transactions but you only want to track the individual customers in order to accurately calculate RPC.
It's easy to see why we want to focus on customers and not transactions. Imagine a customer who refers one friend but comes to your coffee shop every single day for a year. Intuitively, how do we understand the relationship between the customer and the referrals that come from him/her? We clearly would say that the one customer refers one person - we wouldn't say 365 customers refer one person. If we count transactions, we would in effect be saying that it takes 365 of this customer to get a preferred customer - a meaningless and inaccurate statement. Clearly, we can see that it only took us one customer to get that referral - the more accurate approach is to count only customers.
Does the same apply to referrals? Do we count transactions or customers when counting the number of referrals? Clearly, we also count the number of customers - counting transactions would possibly overstate a number of referrals and thereby overstate the RPC metric incorrectly. Again, image one customer refers another and that referred customer comes into your coffees spot every day for a year. Would it be more appropriate to say that one customer was referred or would we say that 365 customers were referred? Clearly, it is more meaningful and correct to note that one customer refers another, not that one customer referred 365 customers.
Digging deeper into the calculation of Revenue Per Customer (RPC)
We touched on the actual calculation above, but let's dig a bit deeper into it in order to really flush out the details. As we said above, there are only two things you'll need in order to calculate Referrals Per Customer (RPC):
Then you simply divided:
(# referrals)/(# non-referrals) = RPC
Make sure to keep in mind that you're dividng referrals by non-referrals (not the other way around). Additionally, it is key that the two numbers your dividing are for the same time period - if you use different time periods for counting the number of referrals and non-referrals, your RPC will inaccurate and incorrect.
Hopefully, you already have the data to be able to get the above numbers. However, if you don't, you'll have to set up a system for collecting customer data and wait a bit (at least 3 months) before you do the calculation. You'll want to wait so that the data is sufficiently representative of what's going on and so the short-term kinks and gyrations are evened out. One year is an even better timeframe - if you start at a shorter timeframe, move to a longer one as more data becomes available. One year is particularly excellent because for most businesses it will allow for a full yearly business cycle (eg. holidays, special sales, varying weather, etc.) to be represented within the dataset you are using.
It's not difficult to start collecting the necessary data to calculate RPC if you currently don't have it - you really only need to tag each customer with whether or not they are a referral and be able to separate out customers from transctions. Separating referrals vs. non-referrals is relatively easy - you or an associate can simply ask at the time of purchase verbally or via a registration form if your business uses them. Making sure transction are separate from unique customers will be a bit more complicated, but is still relatively simple - you'll need to someone keep track of your customers (eg. an MS Excel file) and be able to search within your customer list (eg. Ctrl-F within MS Excel) for the customer when a new transaction occurs. This MS Excel - Ctrl-F is the most basic and primitive approach - far more sophisticated and elegant approaches are possible using both MS Excel or a piece of Customer Relationship Management (CRM) software.
Benefits of knowing your business's ability to generate revenue from customer referrals
Once you know your businesses RPC, you'll have a far better picture of how referrals factor into your business. You will literally be able to understand what percentage of customers are referrals and, thereby, understand what each customer (on average) brings into your business in terms of referrals.
You'll effectively be able to both understand and quantify the additional benefit that is derived from each customer above just the transaction - you'll knw that the transaction amount is only one part of the gain your business receives from each customer. By knowing this, you'll be able to better evaluate marketing - both towards new customers and to existing customers. You'll also be able to better evaluate different approaches to growing sales and revenue - a common dilemma many business owners and managers face is whether to market towards new customers or to focus on getting more referrals.
Additionally, by knowing your RPC, you'll now be able to track your RPC over time - this is incredibly valuable and will allow you to monitor the performance of different strategies and tactics. For example, if you implement a referral bonus where customers get a certain discount for each referral, you'll actually know how effective that program was. You might think that you would already know how effective that program was without knowing you RPC - wouldn't tracking sales and revenue be sufficient? The answer is NO - revenue clearly depends on many thigns (eg. season, tastes, unemployment rate, economic growth, randomness, better salespeople, etc.). By knowing your RPC, you'll be easily able to measure one time period's RPC against another and really know how a new strategy affected the level of referrals derived from each customer.
Most importantly, you'll be able to use your RPC metric in important downstream uses that will further create business intelligence for you - critically useful metrics such as Lifetime Customer Value (LCV) rely on the RCP metric as an input.
Lifetime Customer Value (LCV) is the present value of all gains derived from a customer relationship. LCV is a key metric that big businesses understand (for the most part) and attempt to use in their decision-making process. However, too many small and medium-sized business owners and managers fail to either understand this concept or implement it in their decision making.
Here we'll briefly take you through what Lifetime Customer Value (LCV) means and then walk you through a basic step-by-step guide on how the metric is derived so that you can fully understand how deeply profound and eye-opening the concept can be.
As stated above, Lifetime Customer Value is the present value fo all gains derived from a customer relationship. This sounds simple, but it's not quite as simple as you might think. To really understand what LCV mean's let's break the definition down into its sub-components.
Gains are almost always monetary in final terms, but we say gains instead of money because a lot of the time the final monetary gain comes after multiple non-monetary steps. A simple example would be customer referrals - the referral is a non-monetary gain but a real gain nonetheless because it will end up bringing revenue into your business.
However, often times gains aren't as easy as intermediary steps leading to revenue. Sometimes gains occur due to cost reductions or complex non-monetary benefits. For example, if you're attempting to get your local city government to allow you to put a certain piece of signage up, having more customers come to your business might put some sort of political pressure to get this done. Another example might be the economies of scale that can be achieved by having more customers. All of these complex non-monetary gains must at some point translate into real monetary gains or else they shouldn't be included. Even things such as goodwill, reputation, lax regulatory frameworks, etc. allow for monetary gains down the road.
Now, these complex non-monetary gains are hard to understand and even more difficult to value in terms of dollars. For the vast majority of businesses, it is better to not include them. They will almost surely represent a small portion of your LCV and attempting to introduce them into the LCV calculation will only waste precious resources and potentially cause more harm (in terms of errors) than benefits (in terms of increased accuracy).
So, why would we even mention them if we're not advising including them in the LCV calculus? We mention them because the importance of LCV is beyond the actual number - the profundity of understanding LCV is that you will have a better conception of the nature of your business and your outlook will expand into the longer term. By understand the more nuanced benefits that might accrue to your business (whatever they may be) your overall view of your business - even your underlying emotional and philosophical approach to it - will benefit. Understanding LCV after not knowing it at all is like lifting your head up - while before you were looking at the ground immediate in front of your feet, now you see the entire boulevard ahead of you.
Finally, the gains must be monetary in their final form because we will need to discount them in order to have an accurate LCV. It's almost impossible to discount non-monetary benefits.
2. Customer Relationship
The gains have to come from a customer relationship - meaning someone who has given your business money in exchange for the products or services your business provides.
Of course gains can and likely will come from non-customers - people might refer your business without being a customer because your business isn't selling what they need (eg. a man telling his girlfriends about a new hair salon he's heard or a woman telling her pregnant friend about a store that sells clothes for expecting mothers). However, it is too difficult to capture enough data to be able to effectively understand how much value such people bring. Additionally, the majority of customer value is derived from the actual transactions that take place - that's the heart of LCV and that should form the base of your LCV conclusion.
By using the terms Customer Relationship we are also implying that it's not the immediate interact that is of sole importance - the overall long-term relationship with a customer is key. The main thing to think about here is repeat customers - most businesses have customers coming back two, three, or multiple times. All of these interactions subsequent to the first transaction are clearly part of the value your customer brings your business and should be included in LCV. Of course, subsequent interactions should be discounted (this is addressed below) because money tomorrow is not worth the same as money today (a fundamental principle of economics and finance).
3. Present Value
We've hinted at this above, but it's key to discount your gains in order to properly understand your LCV. For example, if you use your data to see that repeat customers come in every 5 months for repurchases, the payment 30 months from now (the sixth purchase) shouldn't count as much as a payment today dollar for dollar. As in finance and economics, future payments (future gains) must be discounted by the appropriate discount rate in order to come up with the present value of the gain.
Of course, different payments at different times need to be discounted differently - a payment in 6 mo needs to be discounted separately and at a different rate than a payment 6 years from now.
Now, discounting and calculating Net Present Value (NPV) is beyond the scope of this article, but it should be noted that you must discount at the appropriate discount rate - a rate that reflects the inherent riskiness/uncertainty of the future cash flows and the current risk-free rate for that time interval. For example, if a payment is to occur one year from now, you should probably use a discount rate higher than current savings accounts are paying (because that's a rate of a very low-risk cash flow) but something lower than an extremely risky loan (because you're more sure based on your data that the cash flow will come in). This is more complicated than this discussion, however, and any business owner or manager would serve himself/herself as well as the enterprise they are running or managing by taking some time to understand the basics of discounting.
So, we have our recipe for LCV per the above:
- we take all of the monetary gains that will arise from a customer (except those that are complex and difficult to monetize)
- we discount those gains at the apportion discount rate(s)
Going a bit deeper, the most important gain (besides actual money from transactions) is refferals - this accounts for the vast majority of non-transaction gains that most businesses will receive from customers. So, we can simply our formula (while still understanding the broader context from which we are simplifying) to just include referrals - we can literally ignore almost everything else and still come up with a fairly accurate (albeit conservative) LCV. This LCV will be conservative because we're excluding certain gains - it's better to err on the side of conservatism here rather than optimism.
So, we have:
Gains= 1st Transaction + Subsequent Transactions + Refferal Value
But, what exactly does referral value mean? It's not immediately easy to calculate referral value because:
- not every customer will refer people
- not every referral will become a customer
- each customer will usually have a different transaction amount
So, you'll need a bit of customer data to get your Referral Value. You'll need to track how many of your customers are referrals - something many businesses do already. If you don't do this, start it - it's a fundamental part of understanding your business. But just knowing which customers are referrals isn't enough - you need to know how much referrals spend.
You can assume that referrals spend the same amount as the rest of your customers and simply imply onto them the Average Transaction Value (ATV) of your business overall. This is not ideal and can be improved upon with just a little bit of effort. You'll want also get the transaction value of referrals - this is simply done by just recording one other piece of data.
So, now you'll have the number of referrals and the average transaction of refferals. You'll want to use recent data but you'll want to make sure it's a large enough data set - maybe 6 months worth of data at aa minimum. Using that data, you can see how many referrals come in a period of time (let's use a year as an example) and then understand how your referrals relate to your overall customer volume.
Using our 1-year example, say this is what your data shows:
- 1000 customer for an entire year
- 200 referrals
That means 800 (1000 - 200) customers were on-referral. So, 800 on-referrals correlate to 200 referrals for your business. That means:
- every 4 customers is correlated to 1 federal, OR
- every customer is correlated to 1/4 of a referral
Now, you can simply at 0.25*(Average Referral Transaction) to each customer - you now know that a person coming in brings in his money for his purchase PLUS 1/4 of another referral that brings in the amount you calculated for the average referral transaction.
Putting it all together we have,
Gain = Initial Transaction + Subsequent Transactions + Referral Value
To get your LCV, we simply discount this appropriately - a more complicated discussion left for another time.
If you own or run a small or medium sized business and you don't currently track customer, lead, or inquiry data, starting to do so could be magic for your business.
Keeping quality data on customers (and on leads or inquiries - people who contact your business but are yet to take the next step) could prove incredibly powerful in terms of gaining insights about your business and future marketing efforts.
I've spoken to many small business owners both in a professional and social context and I often ask them this question:
"How would you feel if you had the email of every single one of your previous customers available to you right now?"
They usually respond by saying that it would be amazing to have that - the longer they've been in business the more amazing it is. Once small business owner in California that's been around for over a decade remarked that having that sort of data would be like gold - it would allow for excellent marketing opportunities since tens of thousands of happy high-transaction customers have walked through the business's doors over the course of more than 10 years.
Beyond just emails, knowing very simple things like the Zip codes of your customers would add a level of intelligence to your marketing and advertising efforts without which business owners are using primitive methods devoid of any sort of business intelligence that is so easy to acquire in today's environment. Without knowing anything a business owner is shooting in the dark in terms of marketing and advertising. By knowing just the Zip codes of previous customers, marketing strategies could be more finely tuned in order to get better returns on advertising spend. Google, Facebook, and even print ads are all advertising mediums that easily lend to the use of such information.
Any small or medium sized business owner or manager who currently keeps poor data has the ability to transform things today with the use of a computer and a bit of time. The great thing is that it's very easy to start. Although most business owners and managers who haven't started this already have likely procrastinated because they thinks it's too difficult or too complicated for them or their business to implement, that's just false - it's incredibly easy to start for anyone with a computer and a licensed copy of Microsoft Excel.
We recommend using MS Excel (not Google Docs or Apple's Numbers) because Excel is at once easy to use and powerful. Excel is robust enough to handle large sets of data and will allow for various sorts of analysis and manipulation in the future - it's basically the gold standard in terms of entry-level data suites and going for a more simple or easy-to-use suite might be a mistake here.
Here is an example we created to demonstrate how easy it is to start collecting useful data (see image below). We created various simple fields that will be applicable to all businesses:
Date: So you know when the record was taken
Privacy and compliance in collecting customer data
Always make sure to ask for data - don't just input it yourself. Additionally, let the customer know that you're storing it and that you intend to keep it secure and private. Make sure you actually do so by using robust passwords and access protection methods, storing hard drives in safes or secure/locked locations, and refrain from transmitting the data via unsecured connections or to parties who are unaffiliated with your business.
The best rule of thumb is to treat the data as if it was your own or your families' information.
And now, given the rise of cyrptocurrencies and crypto assets to quasi-mainstream financial assets, we're dedicated to providing quality, relevant, and interesting material on cryptocurrencies and cryptoassets. Articles on Bitcoin, Ethereum, Ripple, Cardano, and many more cryptocurrencies and cryptoassets can be found on Pennies and Pounds - all that in addition to a plethora of information on what cryptoassets are, how the entire crypto industry came to be, blockchain/immutable ledger technology, mining, proof of work, proof of stake, and how to prudently invest in crypto if you are so inclined (based on your risk tolerance and ability to withstand the volatility that will come with a crypto portfolio).