How We Know What We Think We Know

In prior installments of this series we took a peek behind CAPE and identified what it has to say about the S&P’s future. This valuation-based metric has statistical and logical basis for telling us some things about the next decade or so of average annual returns.

CAPE suggests the S&P is more likely than not to experience below-trend growth over the next 10 years. But CAPE is just one way of looking at the long-term valuation question to form expectations about the future.

There are lots of other valuation-based indicators, some of which are arguably as useful as CAPE. Price-to-book (PB) is one of them, and the “regular” TTM PE ratio is another. Both have their strengths and shortcomings.

But both also imply the U.S. equities market has “over-trended,” and so we’d anticipate cooler returns in the next several years than in the last several years. And so, because these metrics tell a similar story to CAPE, we mentioned them but didn’t delve into the particulars.

Instead, we took our second run at long-term equities forecasting from out of left field. We infected our minds with an unusual indicator we call HEP that is basically a behavioral-based metric. And while we appreciate HEP’s evident statistical powers, we remain skeptical of its logical underpinnings.

Which is why we couched our discussion of HEP as a means of refining S&P forecasts based on more traditional valuation indicators. Nonetheless, when we put HEP under the microscope like the CDC, we did find additional support for expectations that the next decade of S&P average returns will likely be below trend. And we incorporated that finding into our framework.

All of which is extremely powerful stuff. But also arguably useless garbage.

So we’re gonna make a slight detour. Up next time will be what we can do with our knowledge to obtain superior returns (and I’ll mention a couple of things in this post that relate to that), but for now I feel it’s important to contextualize what we’re up to.

In sum, we’re undertaking a simple exercise in value investing, with a little twist that incorporates some basic behavioral economics. The first stages of this exercise involved touching on pertinent metrics we can use to gauge things numerically and to draw linkage between current-period value and future-period returns, as well as to connect current-period investor behavior with future-period returns.

We’ve seen that very high prices relative to value imply lower returns in subsequent years (and vice versa). And we’ve identified reasons why certain investor behavior trends might be useful in providing an added perspective that refines our value-based math.

But it’s possible we haven’t adequately addressed the promise and pitfalls of this type of exercise. And so now we will.

We Didn’t Know We Didn’t Know It

Because here’s the deal with forecasting like this.

We’re talking stretches of 10 years plus. Even if CAPE, PB, Buffett, HEP, NASA, Napoleon and the CIA all agree the next 10 years of returns will average, say, 3%, that doesn’t mean the annual average for the first 9 years won’t be 30%. Every indicator could be totally right and yet completely misleading.

Moreover, no matter how careful we are, and no matter how many measurements we take, we still run the risk of being proven incorrect. For 10 long years. And we won’t even know we’ve been duped until we’re a decade-plus older!

After all, trends bend and occasionally break. Nobel prizes sometimes seem silly in hindsight. Statistics without valid logic often prove to be nothing more than noisy trifles from the voices in some lonely philosopher’s head. Heck, even Buffett and Napoleon don’t have spotless track records.

“Jeez, FL, you sound all pessimistic about this stuff,” you say. “Are you telling us there’s been no point to this CAPE and HEP and PB acronym orgy?”

Not at all. There’s a great big point with lotsa commas and zeroes behind it.

Nevertheless, I want to make sure everyone’s aware that there will always be noise, and there will always be risk. I’ve said it before and I’ll repeat it aqui: Ain’t no free rides on the market express.



There are sometimes big discounts. And when the price of a ticket aboard that mighty market locomotive deviates significantly from the average, it’s reasonable to conclude the value you’re getting per dollar paid deviates significantly as well. (And “value” in investing is tantamount to future returns performance.)

So the idea is that, by using valuation metrics to form the bedrock of our approach to investing, we can reasonably anticipate greater returns. On average. Over the long run. And we’ll also see that certain behavior metrics can be useful in acquiring greater average returns with lower risk by helping us refine a pure valuation approach.

As a practical matter, that means that, when we peeked behind the curtain of the S&P, we were able to reasonably conclude there’s below-average value right now for the long-term investor. And we’ll soon be able to reasonably conclude there are investing avenues that offer greater value.

Which means the great big point of all this is making more money with our long-term investment dollars with lower downside risk.

We Know We Knew This

But let’s also be clear about how we interpret the findings of long-run forecasting. Because I perceive we humans have a tendency to respond to stuff like this in one of three ways.

The first is to glance at the statistics and pretty graphs and fancy econometrics jargon and promptly conclude on the basis of approximately zero thought: “Yup. Thatsutmgonnado.”

The second way is to draw on the statistics training we received from a part-time wrestling coach in 9th grade and zealously dismiss everything with trite and country-fried erudition: “Cain’t purdict no future.”

And the third way is to go all nukular with numbers and data mine until the Earth is a scorched-out carcass (sorry, engineers, I’m lookin’ at you) and get super-analytical-sure about something that’s not really knowable: “Well, if we combine this metric and the log of that statistic and time-lag by n-3.75 and adjust for wind speed, then we know equities will rise by 2.1863% every 16 months for the next z years.”

And, yup, all three are 100% wrong and absolutely kinda right. Because:

1) There is a role for qualitative reasoning in all this; the statistics can’t tell us everything. Correct. But while we shouldn’t dismiss qualitative reason by just breezily buying into statistics, we shouldn’t ignore the usefulness of good math either. We need both, and can profit from both.

2) The future is intrinsically unknowable and thus not entirely predictable. Okay. But certain long-run trends and tendencies and history-repeating cycles do show evidence of foretelling what comes next on the basis of what’s happened before. And, importantly, in our particular application of different predictive measures, we’re not trying to make sharp and precise numerical estimates; we’re trying to form reasoned directional opinions about relative returns and risks. This is a crucial difference. And it yanks the rug out from under high-waisted clods who aren’t nimble enough to understand the possibility and power of being generally right and still precisely wrong.

3) More data and empirical evidence tend to be better than less. Yup. But only up to a point. We can easily over-soak in numbers and get worse results rather than better – this is a big reason why lots of really smart dudes with tons of data get soaked sometimes. They get so enamored of numbers that they miss the big qualitative stuff. (If you’re not familiar with the dumpster-fire history of Long-Term Capital Management, give it a Google.)

So we practice some prudent caution to temper our use of data and empirical study and analytics. We use qualitative reasoning to investigate what the numbers can and can’t tell us. We remain skeptical of prophesy but smart about using science to our benefit. We don’t buy into every pretty graph, and we don’t haphazardly dismiss good data.

And, ultimately, we adhere to that simple and superficial and country-fried and slide-rule-approved maxim of value investing: Buy low and (sometimes) sell high (but mostly just hold).

I Knew It!

And that’s what we’re doing here.

Consider: If we’re not in the business of finding good value with our investments, we’re damn near gambling. We oughtta just hit the slots in Vegas instead. At least we’d get watery gin-and-tonics for free on the floor of the Golden Nugget.

How do we find good value?

Well, we do what we’ve done these last couple posts. We take a look at valuation indicators that are ever-so-slightly more useful than swinging wild ass guesses. And we form reasoned opinions about what those indicators imply. And then (as we’ll do next) we make determinations about the sorts of actions merited by those opinions and indicators.

There’s no mystery to it. And no sorcery. With a little bit of training and discipline, there’s nothing to all this that couldn’t be done in an afternoon by anybody with a dial-up connection and a pencil.

But, likewise, there’s no mystery or sorcery or special intellect required to know we shouldn’t spend more money than we earn. That we shouldn’t eat way more calories than we burn. That we should exercise regularly. That we shouldn’t smoke cigarettes. That we shouldn’t do all sorts of amazingly stupid crap.

And yet, most American households barely have a positive savings rate and/or have serious debt, don’t have anywhere close to sufficient savings for retirement, have members who are overweight or obese or diabetic or smokers, and have a serious addiction to doing all sorts of amazingly stupid crap.

So: Just about anybody could undertake this exercise if they wanted. But they don’t. And even if they did, it’s not likely they’d do what they should. (This includes professional money and investing dudes – who often have other problems to deal with like incentive misalignment, temporal constraints and scale issues.)

Which is another way of saying that, just because there’s no good reason most people don’t do the smart thing with their money, it doesn’t follow that there’s no way to benefit by doing that smart thing – even if it’s obvious. Most people, regardless of whether they know the right answer about stuff, fail to do anything about it for a zillion reasons that defy reason itself. Or they do exactly the wrong things for no reason whatsoever.

All of which is to say there’s profound power in undertaking the sort of exercise we’re in the midst of here. But it’s not absolute power. It’s not unequivocally infallible. It’s not, over every interval and in comparison with every alternative approach, always going to be superior in hindsight.

But it will be better, on average, over the long term. And it’s the only approach to financial decision-making that always makes prospective sense.

After all, finding good value on the basis of reasoned and qualitatively-textured economic reasoning and data-focused research is a tried and true approach to obtaining supra-normal returns over the long run…whether it’s in business or in the markets.

Think Graham. Think Buffett. This is the program. It’s as staid and starched as a plateful of white rice. Ain’t no 1,000% overnight returns with stuff like this. Because, unlike the dumb money on the casino floor in off-strip Vegas, this ain’t gambling.

This is business decision-making oneohnotquiteone.

You Know It

That’s why we like valuation-based indicators like CAPE and PB. They’re the heart and soul of finding value. And of telling us when value is more likely than not to be missing.

But, as we know, peeps are hardly the coldly calculating Spocks that examine value and nothing else, that always make the rational choice, that never do the dumb thing. Peeps be peeps that behave like humans – stupidly, brilliantly and often very predictably.

Which is why behavioral indicators like HEP can have a useful role. And, as we’ll see, there’s compelling empirical support for behavioral metrics as insightful complements to traditional valuation indicators in identifying investment opportunities with above-average return and below-average risk profiles.

And so we’ll use them.

Next time, that is.

Where we go from here is where the grass is greener, the air clearer, the risk-adjusted returns greater and the high-waisted clods gratefully outta sight.

We go where the reasoned and judicious use of meaningful data and qualitative reasoning take us.

We go where most don’t dare to (or just don’t care to).

And that’s exactly why where we’re going will make us money and make us look like we knew a whole lot about the future, even if all we really know about is the past.

Luchadores, this post was not part of the original plan for this series. But I felt it necessary at this juncture to make sure we’re all on the same page about exactly what is – and isn’t – being said and done here. Holla with thoughts about the right role for data and qualitative reasoning in rendering forecasts – I’d be interested to know your thoughts on the right blend in making financial decisions.


Libre Your Mind