By Dr. Jim Dahle, WCI Founder
Selecting investments solely by looking at past performance is the equivalent of driving while looking in the rear-view mirror. Sure, it can be done, but it probably isn't going to lead to optimal results—and it might even lead to a spectacular crash.
This is known as “performance chasing” and is a well-known behavioral finance error that leads to repeatedly buying high and selling low. Yet most beginner investors do precisely this.
An example was posted recently in the WCI Facebook Group.
I was pleased to see that people in the group were very kind, and they lovingly helped this investor to see the error of their ways while providing lots of great education about returns, yields, and investment selection. But I thought this was too important of a topic to allow it to be buried in the depths of Facebook. This sort of thing happens all the time when people are asked to start investing. When does that occur? It occurs when they sign up for a 401(k) for the first time. They're shown a dozen—or even worse, 50—different investments, and they are asked to select one or more in which to invest their hard-earned dollars.
For example, when I go into my partnership 401(k) to select investments, this is the screen I see:
That's it. Just the names of the funds.
Now, this is a pretty darn good list of funds. If your list looks like this, you have nothing to complain about.
But most people don't have this sort of a list. And in fact, if I click on one of the funds, it takes me to the Morningstar page for the fund, which has all kinds of useful information on it about the fund holdings, fees, and past performance. (Incidentally, that's the order of information to look at when selecting funds.) But if you stay on the 401(k) website, all you can really learn about the funds is the current share price (useless information when selecting funds) and the past returns (almost useless information when selecting funds).
The performance looks like this: columns of returns including 1 month, 3 months, YTD, 1 year, 3 years, 5 years, and 10 years.
That's the page most people see when they're choosing funds for their 401(k). If that's all the information you have (or look at), you're, of course, just going to pick the things that have the highest past returns.
The Problems with Past Investment Returns
There are a number of issues that occur when you're looking at past returns. The first one is that many sources don't actually report them properly. Most investors don't even know how to calculate their own returns, and a surprising number of investment managers don't do it properly either. But many news sites aren't even close to doing it properly. For example, they just report the change in share price and ignore all of the dividends. For example, if you Google “VTI” and look at the chart . . .
. . . You'll see 17% returns. But if you go to the Vanguard website, you'll see it's reporting 19.2% returns for almost the same one-year time period. Why the difference? Well, there are actually two differences. The first is the Google chart was captured on the afternoon of February 2, and the Vanguard return was reported as of the end of the day on January 31. The market went up a fair amount in that day and a half, and maybe it didn't go up that much from January 31-February 2 in 2023.
But the main difference is simply that the Google chart excludes dividends. If you exclude dividends, as so many news sources are apt to do, stock returns look significantly worse than they actually were, especially in the long run. This is precisely what the Facebook poster was doing. A chart was posted that only showed the change in the share price of the bond fund. That's just silly when you understand that almost all of the return of a bond fund is from the income of the fund. Even when things are going pretty well, the price of the fund doesn't go up much. In fact, the return on that particular fund over the last five years has been positive (0.85% per year on the day I'm writing this), not negative as the chart would suggest.
However, once you know you're actually looking at the true returns, there is still a major issue with using them to decide how to invest. That problem is that you don't have any plutonium to put in your flux capacitor.
You can't go back and get those returns. They're gone. And they're really not very predictive of future returns, especially short-term returns. Don't believe me? Spend a little time with the Callan Periodic Table of Investment Returns.
Each color is a different type of investment. Notice how there is a different color at the top of the chart every year. If you just bought what did the best last year, you're going to have pretty rotten long-term returns.
There is likely a weak inverse correlation between short-term past returns and short-term future returns. The reason for this is obvious to the sophisticated investor. Short-term poor returns generally mean the price of the investment has declined. That means you're buying essentially the same thing for a cheaper price. Just like you'd love to buy a hamburger and gasoline and that cute little skirt on sale, you should love to buy stocks, bonds, and real estate on sale. All else being equal,
-
- When the price of high-quality bonds goes down, the yield goes up, and the best predictor of future bond returns is the current yield.
- When the price of stocks goes down, the price-to-earnings (P/E) ratio falls, and you're buying more company earnings with the same amount of money.
- When the price of real estate goes down, the capitalization rate goes up, and even without any future price appreciation, your expected future return has gone up.
For any reasonable long-term investment, a lower price (meaning recent poor returns) typically increases expected future returns. There's more to it than just “buying the dip” or buying the worst-performing investment you can find, but you certainly shouldn't be scared off just because the investment has had poor returns for the last month, year, or even five years.
People are inappropriately reassured when they look at a five-year return chart and see good things and inappropriately scared when they look at a five-year return chart and see bad things. You've got to dive deeper. Why do those charts look like they do? What was different five years ago than today? If I were buying a total bond market fund today, I'd expect returns over the next five years to be something like 4.3% per year. But there can be a bit of a range. It could easily be 7% a year. It could also easily be 1% a year. If interest rates rise, that will likely result in lower returns than 4.3% (the current yield). If they fall, that will likely result in higher returns than 4.3%.
More information here:
The Nuts and Bolts of Investing
10 Ways to Console Yourself When Losing Money in the Markets
Know What You're Buying
When it comes to investing, you need to understand what you are investing in and how it works in different economic climates. This will eliminate surprises and make it easier for you to stay the course with your long-term investing plan. Past performance numbers (especially very long-term past performance numbers like 20-30+ years) can be a useful part of the evaluation of an investment, but it's not nearly as useful as novice investors think they are.
Physicians train for years to learn about medicine. But financial literacy was not part of the curriculum. That’s where The White Coat Investor comes in—by offering tons of entry-level information to get you started on the right path. We have a FREE email series called WCI 101 that reviews the basics in bite-sized chunks. You can check out our Start Here page to learn all about personal finance for doctors. And you can peruse our Frequently Asked Questions to get even more info. It’s easy to feel overwhelmed when learning about finance. WCI is here to help!
What do you think? Why do so many people invest while looking in the rear-view mirror? What can we tell them that will help them to stop?
“Most investors don’t even know how to calculate their own returns, and a surprising number of investment managers don’t do it properly either.”
Even though I’m a reasonably sophisticated investor, I confess I’ve been struggling with this as I do more tax-loss harvesting and donating appreciated assets. Imagine a hypothetical portfolio containing only two equity assets, an S&P500 index fund (S) and a total stock market index fund (T).
When the market drops, I sell S and buy T
When the market rises, I donate T to my DAF and immediately buy more T with new money
When the market drops, I sell T and buy S
When the market rises, I donate S to my DAF and immediately buy more S with new money
When the market drops, I sell S and buy T… etc.
How do I calculate my returns, accounting for the tax losses and charitable deductions?
@UNHWI,
The easy answer is not to try and reinvent the wheel as the Bogleheads already have a pretty good “returns spreasheet” that I believe is located here:
https://www.bogleheads.org/wiki/Calculating_personal_returns
If you don’t want to go to all the trouble of loading all your old data and already have your own spreadsheet with monthly totals of your accounts all you have to add to this are totals for withdrawals or contributions and then use the formula below:
[EV – BV – Adds] / [BV + 1/2 Adds]
where;
EV= ending value
BV= beginning value
Adds are contributions (+), or withdrawals (negative adds)
I keep my portfolio totals monthly, which I think is what the Boglehead spreadsheet supports.
I think the above method which assumes you made the addition or withdrawal at the midpoint in the period is accurate enough for most portfolios. Your broker does something similar but calculates the gains on a daily basis before adding in or subtracting deposits or withdrawals. Obviously, the larger the size of adds or withdrawals compared to the total will affect the results.
The best way to calculate your own returns is to use the XIRR spreadsheet function in Excel or Google Sheets.
https://www.whitecoatinvestor.com/how-to-calculate-your-return-the-excel-xirr-function/
Thanks WCI and FinancialDave.
The limitation with these spreadsheets is that they don’t capture the tax benefits of tax-loss harvesting and donating appreciated assets to offset income. In the specific example I gave above, the returns calculated using XIRR or the Boglehead spreadsheet would be lower than the returns of the S&P500, but the after-tax returns would be higher, at least for higher tax brackets. How do you calculate these differences?
I never bothered calculating after-tax returns. I guess add up whatever tax bill you have for your investments and subtract it as a withdrawal in the XIRR calculation.
@UHNWI,
Here is one way to handle your dilemma. It is easier if all your taxable accounts are consolidated at one broker, with one account, but the idea is the same for multiple broker accounts.
Your spreadsheet handles everything that goes in or goes out of your taxable account. This includes the taxes paid due to your taxable account, which generally can be pulled from your 1099B for this account. Granted this information comes in a different tax year, but if you do your own taxes or at least can use tax software, you can backout essentially the 1099B that applies to see the net effect of everything going on in the taxable account.
I handle this for the complete picture of all my accounts by “stacking” my taxes in such a manner that each taxable income source stands on its own. Since I am retired it looks like the following:
1. 10% tax is withheld from our Social Security – which covers this tax.
2. I have a pension that pretty much matches my Standard Deduction, so no tax withheld from it.
3. All other money spent then comes from either TIRA or taxable accounts. In my case taxable causes very little effect, but if it did, I could pull money from taxable to pay that tax. TIRA withdrawals stand on their own with 12% tax withheld and sent to IRS from the broker.
In your case it would be a worthwhile experiment to use some 2024 tax software, if you haven’t already, get the total picture of your taxes for 2024 and then pull out everything associated with the taxable and see how your net tax changes. This number is then a “withdrawal” from your spreadsheet to cover the taxes, depending on where you paid them from and whether the net is plus or minus to your net taxes.
Thanks FinancialDave for the thoughtful comments. Unfortunately I have income from multiple sources, some of which is reported on K-1s, which I’ve always found difficult to interpret.
Also, in contrast to my simple example with two funds, I’m actually doing tax-loss harvesting and donating appreciated shares using direct indexing, so several hundred positions.
But I like your suggestion of looking at cumulative tax losses for the year as well as charitable donations and using these to crudely calculate how much extra tax I would have paid.
@UHNWI,
I’m not sure it is as hard as you make it out to be. Dividends, income from K-1s, etc are not an input to the spreadsheet, as are any of your trades.
If you look at the spreadsheet inputs say for the Boglehead’s one, each monthly Portfolio has only 3 entries; date, month end total, and adds to the portfolio from outside the broker( plus for deposits and minus for withdrawals.) Nothing that happens inside the account, such as buys, sells, dividends, etc matters. It should take less than 5-10 minutes per month for a half dozen different portfolios.
What is a little more work is sorting out the taxes at the end of the year, but after doing it once you will have an idea of how much that taxable account is costing you every year and you may decide like I did, that I’m not going to spend it anyway, so I just put it in tax efficient investments and “let it ride.” I suggest you give it a try – seeing is believing!
One theory why people look in the rear view mirror:
We’re creatures of habit. If we like a restaurant, we go back there. Even more, if we liked the dish we ordered, we have it again. We base our current decisions on prior history.
When it comes to investing, we’re wired to do the same thing. We chase prior returns because in our minds, we like those prior returns and can’t comprehend why things might be different this time around. Just like the Italian Beef was good last time, it’ll be good this time.
How do we adapt our mindset when it comes to investing? Trust the fundamentals, such as paying attention to fees, like you mentioned above. Invest early and often. Ignore the noise.
Spend your energy and time enjoying good meals instead. Maybe like an Italian Beef.
Thanks,
Matt
@Think….,
When I’m driving down the road, especially a multi-lane expressway, I am constantly checking both side mirrors and the rear-view mirror as I drive. When an unforeseen situation occurs, you may need to react by moving to a different lane in a hurry. However, if you are on a multi-lane highway in heavy traffic, it is best not to “lane hop”, which could be related to jumping in and out of stocks based on current performance. I have found even on the freeway in heavy traffic, I sometimes get where I am going sooner by staying in the “slow lane.” You see everyone else is in a hurry and jumping in and out, slowing things down in the faster lanes!
Well said, Dave!
In your honor, I’m going to drive home during Chicago rush hour in the slow lane tonight!
Matt
@WCI,
“When it comes to investing, you need to understand what you are investing in and how it works in different economic climates. ”
Maybe I missed something in this article, but if I need to use something other than past performance, what would that be?
Use for what?
@WCI,
When you pick an asset to be in your portfolio (let’s stick with equities here) you have 10’s of thousands of choices from stocks to ETFs and mutual funds. If you “can’t invest by looking in the rear-view mirror” ala past performance, then the only other kind of performance metrics would be called “future performance”.
Think about why you can look out the front window of your car to get where you are going. That is because someone in the past has mapped out where the road goes, otherwise you are just driving around “hoping” for a good outcome.
Now granted, part of your title suggests stocks are not going to follow a predetermined path, such as a known road, which is entirely true, yet still everything we know about stocks comes from past performance, those who use that past performance may still not get to where they want to go, but it is all we have, other than future performance.
I think the issue at hand is not whether it’s okay to use past performance or not in making decisions. The issue is HOW you use that information. What I am warning about in this post is people who see that bond returns were crummy for the last 3 years including 2022 and decide based primarily on that fact that they should not invest in bonds.
Nobody is saying “Don’t look at historical, especially long term historical, returns of the various asset classes when building a portfolio.” That’s a straw man that you have easily knocked down as expected.
@WCI,
The problem is your title, it is not justified by your article, in any way that I can tell. You make a very broad statement about past performance, yet you are the one with a weak argument that just because some people draw the “wrong” conclusion from past performance of bond funds over a short time period, that past performance can’t be trusted! This sounds in itself like your own strawman argument.
In fact, after re-reading the article a few times, most everything you are talking about has to do with past performance and “how” to interpret it, but in the area of bonds you don’t even get into HOW to interpret the past performance. Bond performance has to be interpreted based on the duration of the assets you are considering, for instance; a bond fund such as IEI with a duration of about 4 years and a yield of about 4% can be reasonably expected to return to you in 4 years that 4% annualized return, and maybe a little better if the interest is reinvested. To try and interpret bonds in any other light is just a “failure” of perception from most who understand how bonds work – “so I have been told!”
A failure to interpret “past performance” accurately does not negate the fact that others may use that performance to draw the correct conclusion – which may in fact be “don’t invest here.” Or in the case of the example IEI bond ETF, you need to hold the fund for 4 years if you expect to get something near the 4% that is suggested right now.
The information about likely future returns from the bond fund is coming from the current yield and duration, NOT the past performance. That’s my point.
@WCI,
Yes, that is partially true, which I will explain in a bit, but first let’s get to what I should have led with. Your title IMO was originally meant to caution investors from investing in the “stock market” by looking at prior results. It originated as a requirement by regulatory bodies like the U.S. Securities and Exchange Commission (SEC) to protect investors from being misled by historical returns when evaluating investment products like mutual funds or stocks. It has now been so overused and attached to anything including your laundry soap and many medical drugs as well!
Bonds funds are somewhat a unique product not really related to equities and their past returns as I explained above. If we are talking about a simple bond fund, made up only of bonds you have to look internally at what it is made up of to determine if past performance is a contributor, which it absolutely is.
Start by looking at buying one individual bond with a fixed coupon and non-callable. When it is first bought there is no past performance and IF you hold it to maturity you will get exactly what you expect, however if you sell it at some point before maturity, its price may have dropped by half and this “past performance” would be due mostly to a change in interest rates over the time you owned it (past performance.) So, one thing you can say about bonds is you “could” avoid past performance affecting you, barring Internal “defaults, by holding your bond to maturity.
Now have someone put together a portfolio of the above bonds and call it a long-term treasury like TLT. Each of the internal bonds performs just like explained above. If there were only two internal bonds, purchased on the same date you did, it wouldn’t be too hard to determine exactly what you would get if they both had the same maturity date and you held it to that date, In the real world that is not the case, as of today TLT has 42 holdings. To demystify what you are buying there are a number of “past performance” parameters that can at least give you a good indication of your results by boiling this TLT into a single investment with an effective duration of about 15 years with a current yield around 4%. Meaning that if you hold this investment for 15 years and reinvest the dividends you will get something close to a 4% return, regardless of where interest rates go (past performance) in between. If you sell early, you are very much subject to the past performance of interest rates from the time you bought until the time you sell. This is what “most” bond investors do not understand and why your investors are asking “what is the point” of investing in bonds, if they aren’t predictable? The problem is quite a common one, when investors see bonds as something to trade and don’t really understand how the past performance of interest rates can affect their investment if not held for the “duration.”
In 2020 at least one money manager was advising people to buy TLT at a time when it was selling for $159, with an effective duration of 13 years and a yield of around 2%, thinking incorrectly that 2% would look pretty good compared to a market going “south.” No chance to lose money he said. Well yes, if you want to hold that investment for around 13 years. Now rounding slightly that share(s) bought at 160 are now worth about half and on that basis, he is getting paid around 4%, due entirely to the past performance of the interest rates, that puts him now at around 1/3 of the way to paying off the $160 shares he bought in 2020 and earning 2% for 13 years.
Turns out both current yield and duration are something that come from “past performance” of interest rates and internal bond maturity dates and changes continually with time. That past performance from the time of your purchase does determine whether you can sell your investment for a gain or not prior to the “duration,” much like selling an individual bond prior to its maturity.