Averages, Beverages, And The Benefits Of Baseline Returns

Advanced, Theory

If you’re one of the millions of people launching into the gift-buying spree this holiday season, then there’s a good chance you’re inundated with numbers right now.  From performance statistics and customer ratings to price points and discount percentages, smart shoppers are in a constant search for the best bang for the buck.  Even if you’re not the type of impulse shopper that particularly enjoys the experience of browsing the aisles for just the right gift, experienced marketing professionals have learned that maximization of value is a powerful motivator that can pull even the strongest introverts into crowded stores against their better judgement.  Why do you think they spend so much money on Black Friday ads touting record-breaking deals?

After a lifetime of this type of shopping conditioning, it’s no wonder that this same maximization mindset might bleed into other decisions as well.  Like, for example, what asset allocation you might choose for your life savings.  So the same highly intelligent shoppers often create very similar lists of investing options sorted by the most common performance metric available — average return.  Just like how you might seek the best resolution for a new computer monitor or the highest customer rating for a new toy, you surely want the highest average return for your hard-earned savings.  Right?  

Unfortunately it’s not that simple.  Contrary to your data-driven instincts, averages lie.  So take a break from your holiday shopping, find a comfortable chair with your favorite drink, and let’s talk about how averages distort our thinking.

Imagine for a moment that you’re shopping for a bottle of your favorite adult beverage.  Beer is my weapon of choice, but any good drink will suffice.  Whether you go by your own past experience or the group ratings of others, your goal is generally to find the best option at a certain price or the best price for your favorite drink.  Thanks to the miracle of modern bottling methods where a bad batch is exceedingly rare, you have a pretty good idea of what to expect with the average drink and can look around to find a good deal.

Picture taking that first sip only to discover that the drink is a little flat.  You do some research and find that even though the sample everyone is offered straight from the source is an award winner, the smallest exposure to the real world immediately dulls the taste.  While it might not meet expectations and makes you think twice about average reviews based on an idealized and unrealistic product straight from the tap, you understand that it’s a fact of the industry that everyone deals with and drink on.

Next, assume that the consistency you take for granted does not exist and the taste of any one bottle is wildly variable.  The average bottle might be quite good, but there’s a one in six chance that the one you purchase will taste like pure vinegar.  Talk about a nerve-racking six pack!  Let’s say the brand next to it at the store has a slightly lower average rating but can guarantee the quality of the product for the exact same price.  Which one would you choose?  Would you still go with the better average bottle or choose the one guaranteed to provide a satisfactory experience?

Finally, what if I told you that the occasional bad drink fundamentally changes your body chemistry and permanently diminishes the flavor of every drink you consume from that point forward until you get lucky enough with a string of particularly good bottles to detoxify the system?  Is it really worth it? 

Add up all of those bizarre product behaviors and it’s probably enough to drive a lot of people to permanent sobriety.  Clearly the average taste is not at all the right metric to describe the up-and-down experience of consuming such a volatile beverage with questionable initial ratings, uncertain product quality, and long-term consequences to every sip.  

Luckily drinks don’t work like that.  But believe it or not, that strange vision of a beverage-based game of chance is exactly how investment returns behave.  Between inflation that immediately flattens the advertised purchasing power from the pristine nominal return, sequence-of-return uncertainty that deeply sours the outcome for unlucky investors, and compound returns that force the bad taste to linger for years and affect your account balances far into the future, investments play by a different set of rules.  Those average returns that you naturally believe are a reliable measure of quality are largely meaningless for scores of investors.

I believe new measures that capture the broad range of outcomes (including frank discussions of less desirable scenarios) are necessary to fully educate prospective customers before they partake. So what might such a new & improved measure look like?  Recapping the above example, it would need to do the following:

  1. Always account for how inflation erodes purchasing power
  2. Measure the true compound growth that people actually experience
  3. Capture the idea of returns uncertainty in a simple and conservative way

I personally don’t care for nominal average returns at all as they accomplish none of those three goals.  Instead, I prefer what I like to call the baseline return.  You’ll find references to it in things like the Long Term Returns chart and the Portfolio Finder, and it’s my favorite metric for setting realistic, conservative, and data-driven portfolio expectations without trying to predict the future.  

The technical definition of the baseline return is that it’s the 15th percentile inflation-adjusted compound annual growth rate (CAGR) for a given investing duration looking at every start date we have access to.  But practically speaking, the baseline return is just a reasonably conservative real-world annual return that accounts for inflation, uncertainty, and compound returns while excluding the worst outliers.  Or if you’re a visual person like I am, you can find it in the Long Term Returns chart within the various Portfolio tools.  This chart displays the full range of inflation-adjusted (aka “real”) CAGRs starting in every year since 1970 based on how long you stayed invested.  



See that pink diamond marking the border between the bottom 15% of all 10-year returns and the middle 70%?  That’s the 10-year baseline return for this particular portfolio.  While the median return may look very tempting, it’s better to be realistic and acknowledge that half of all investors earned less than that.  Using the baseline return will set you up for far more realistic investing expectations and allow you to build a reliable savings plan to meet your important financial goals no matter what the future holds.  

So how much of a difference does it make to use the baseline return over the average return when selecting a portfolio?  You might normally expect high average returns to correlate to high baseline returns, but not all portfolios are created equal and variations in the underlying uncertainty of each asset allocation can really affect the numbers.  For example, here’s a comparison chart sorting every portfolio on the site by its long-term 15-year baseline return while also displaying its long-term average since 1970.  Note that both numbers account for inflation in this example, so the differences all come down to uncertainty and the effects of compound returns.

First take a look at the Total Stock Market on the right.  This has long been a popular choice among aggressive investors seeking the highest average returns, and the data agrees that its average was highest of all the options by quite a bit.  But once you account for the effects of uncertainty and compound returns, the baseline return was dead last!  Dependability is certainly not a hallmark of the stock market, and investors are not at all guaranteed that high average in their own investment accounts.  How would you feel if you expected the advertised real return of 7.9% a year and 15 years later you look back and realize you only experienced 3.3% with your own money?  I’m guessing you wouldn’t be too happy.  
Or what if you expected the nominal 11.2%?  Based on the reactions I see every day to people freaking out about just a few months of under-performance, severe unhappiness would likely be the best case scenario.

Next, look at the portfolio with the highest baseline return — the Pinwheel Portfolio.  It was no slouch in average return either, as it finished second in that metric.  But with a baseline return of 5.8%, I imagine even the unluckiest investors would have been reasonably satisfied.  Wise investing is about so much more than maximizing the average return, and the benefit of true economic diversification is that it greatly reduces portfolio uncertainty while still providing great performance.  

And finally, take a look at just how random the average return looks when compared to the baseline return.  Some portfolios are simply more consistent than others, and average return is a poor predictor of baseline performance.  Options like the Rick Ferri Core Four had high averages and middling baselines, while others like the Larry Portfolio had low averages but pretty good baselines.  To find the right portfolio for your personal needs, you really have to learn to set aside the average return and begin to look deeper.  There’s so much more to learn about performance beneath the surface.

That search for truly meaningful and useful data is why I personally study asset allocation and is what Portfolio Charts is all about.  It’s no accident that the top two baseline return performers are my own ideas, as reducing portfolio uncertainty while still providing highly desirable outcomes even when markets don’t cooperate is one of the core investing insights I truly believe in and want to share.  But I also don’t claim a monopoly on productive investing styles, and the site is full of interesting portfolios created by smart people for all types of investors. 

So don’t just take my word for it.  Pour another round and explore for yourself.  Averages may lie, but good decisions do require good data and there’s plenty to go around.