Fix It Friday Ep. 32 - Patterns Aren’t Predictions: The Extrapolation Mistake

• Past market performance is not a reliable predictor of future returns.

• Mean reversion is a long-term concept—not a short-term forecasting tool.

• Behavioral biases often tempt investors to abandon disciplined investing.

• Successful investing depends on consistency and long-term compounding, not market predictions. 

Disclaimer: [00:00:00] The following podcast by Fusion Family Wealth LLC, Fusion, is intended for general information purposes only. No [00:00:05] portion of the podcast serves as the receipt of, or as a substitute for, personalized investment advice from Fusion or any other investment professional of your choosing. [00:00:10] Please see additional important disclosure at the end of this podcast.

A copy of Fusion's current written disclosure brochure discussing our advisory [00:00:15] services and fees is available upon request or at www.fusionfamilywealth.com.

Jonathan Blau: Welcome to another [00:00:20] episode of the Crazy Wealthy podcast, Fix It Friday edition. So today, I'm gonna talk [00:00:25] about patterns, and particularly the way we perceive patterns and how that [00:00:30] perception can cause us to make a very critical investment mistake.

And the title of today's [00:00:35] Fix It Friday is Patterns Aren't Predictions: The Extrapolation Mistake.[00:00:40]

Voiceover: Welcome to the [00:00:45] Crazy Wealthy Podcast with your host, Jonathan Blau. Whether you're just starting [00:00:50] out or are an experienced investor, join Jonathan as he seeks to [00:00:55] illuminate and demystify the complexities of making consistently rational financial [00:01:00] decisions under conditions of uncertainty. He'll chat with professionals from the advice [00:01:05] world, entrepreneurs, executives, and more to share fresh perspectives [00:01:10] on making sound decisions that maximize your wealth.

And now, here's your [00:01:15] host.

Jonathan Blau: So let's start with a [00:01:20] question. If the stock market has gone up about fifteen percent a year in the past, uh, [00:01:25] six years or so, does that mean that the next six years are likely to be bad? [00:01:30] That's what a very smart and successful client asked me recently, and it sounds [00:01:35] logical. But today's Fix It Friday is about why that conclusion, while intuitive, is [00:01:40] actually one of the most dangerous mistakes an investor can make, because it's based on something [00:01:45] called extrapolation, and extrapolation is a long-term killer of [00:01:50] compounding.

So here's what's happening. Someone looks at the last five or six years, sees very [00:01:55] strong returns, and concludes, "Hey, returns have been really high, above average, so they must [00:02:00] now be due to go on a streak that's below average going forward. Let me change my [00:02:05] portfolio accordingly." And that sounds like analysis, but it's not.

It's extrapolation. [00:02:10] And it doesn't just show up in one direction. If we flip it around, you'll hear the mirror image [00:02:15] just as often. "We've lagged in our returns for the last five years in the markets, so now we're [00:02:20] probably due for a big bounce. Let me change my portfolio in anticipation [00:02:25] of outsized returns."

Same insti-- st- same instinct, same trap, just [00:02:30] pointed the other direction Extrapolation is when we take a real pattern we see [00:02:35] in past returns and make the leap that it implies a predictable [00:02:40] future path, and that's the key, key word is predictable. Investors tend to do this in [00:02:45] two common ways. The first is after a strong period of recent returns, people [00:02:50] say, "We're ahead of average, so now we must be due for lower returns."

The second is after weaker [00:02:55] long-term returns, people say, "We're behind average, so we must be due for higher returns to catch up." [00:03:00] Both of those ideas are rooted in something real, and it's called mean reversion. [00:03:05] Here's a simple way to think about it. Mean reversion is the idea that if returns come in below [00:03:10] or above their long-term average for a period of time, at some point, returns [00:03:15] would have to come in higher or lower to bring that long-term average back in line.

[00:03:20] So I'll give you an example. If returns were five percent for a period of time and the long-term [00:03:25] average has been 10%, you would need a period of higher returns at some point to [00:03:30] get back to that long-term average of 10%. That's a useful long-term observation. [00:03:35] But here's where things go wrong. Investors take that concept and turn it into a short-term [00:03:40] forecast.

They assume, "Hey, we're above average, so now we must go below," [00:03:45] or, "We're below average, so now we must go above." But mean reversion doesn't work on our [00:03:50] timeline. It doesn't tell us when it happens. It doesn't tell us how it happens, [00:03:55] meaning when the returns of the long-term average is gonna be occurring and, and how it's [00:04:00] gonna happen.

So mean reversion tells us something about long-term outcomes. It tells us [00:04:05] nothing about what happens next, and that's where investors make the biggest mistake. [00:04:10] And this is where a few behavioral biases sneak in. So one is called framing [00:04:15] bias. It's when we cherry-pick a timeframe to make it look like the facts currently [00:04:20] support the narrative we already believe.

Start the clock in 2020 and [00:04:25] recent returns look unsustainably high, and we feel like we must be due [00:04:30] for an immediate reversion period where, where returns are gonna be lower than they have been. Start that [00:04:35] clock in 2000 and the story flips completely, and I'll explain why [00:04:40] in another couple of minutes Recency bias is where we overweight [00:04:45] what just happened, and we let it drive longer-term expectations and [00:04:50] decisions.

The last twelve months feel more real to us than the last twelve [00:04:55] years, even though they matter far less to a twenty-year plan. So we look at [00:05:00] recent periods and extrapolate from that into the future. Denominator [00:05:05] neglect, this is an important one. We treat a short window of time, meaning the [00:05:10] last five years, like it's meaningful.

Five or six years feels like a long stretch to live through. [00:05:15] It's a rounding error in a thirty-year or even multi-generational investment horizon. [00:05:20] Another example I like to use of denominator neglect is when you hear the media try to scare [00:05:25] us, which is one of their jobs, so we can click on their stories and increase their ad revenues.

Uh, the-- [00:05:30] today, they'll say that the Dow Jones is down a thousand points, and that sounds awfully scary. [00:05:35] But the denominator, which is the current level of the Dow, not just the numerator, the [00:05:40] thousand points, is fifty-two thousand. So now it's a thousand points down over [00:05:45] the denominator of fifty-two thousand.

That's less than a two percent decline. Fairly [00:05:50] common in the markets But if you take a look back in April of [00:05:55] twenty twenty when the Dow Jones average was twenty-one thousand, uh, the same thousand-point [00:06:00] decline is about five percent or more, so that's much more meaningful. By neglecting a [00:06:05] denominator, in this case, the level of the Dow at twenty-one thousand in twenty twenty [00:06:10] or at fifty-two thousand today, that same thousand-point scare tactic that the media [00:06:15] uses by intentionally neglecting the denominator where the Dow is can be awfully scary [00:06:20] when it shouldn't be.

So a two percent decline versus a five percent decline is very different, [00:06:25] but a thousand points sounds scary when they don't include the denominator, no matter what the [00:06:30] percentage is. So den-denominator neglect is, is, is a big, uh, behavioral, um, [00:06:35] foible that we need to avoid. Mean reversion misuse, turning a long-term tendency [00:06:40] into a short-term prediction.

It's the difference between knowing winter eventually comes and [00:06:45] knowing whether it starts next Tuesday or next Thursday. Put those together, and we [00:06:50] don't just observe patterns. We start believing those patterns can tell us what comes next [00:06:55] or can predict the future. So let's pressure test this with real numbers.

Look [00:07:00] across rolling multi-year periods for the market going back decades. Some strong [00:07:05] stretches were followed by more strong years. Some weak stretches were followed by more [00:07:10] weak years. Others reversed entirely. There's no reliable rule that says strong [00:07:15] periods have to be followed by weak ones or the other way around.

The only consistent [00:07:20] finding is that recent performance by itself tells us nothing about what will [00:07:25] happen next. That's not a hunch. That's just what the data shows. So let's change the frame [00:07:30] now, as I promised. Let's look back, and instead of starting in twenty twenty, let's go [00:07:35] back to the end of nineteen ninety-nine.

Now we're looking at a much broader period that goes back [00:07:40] about twenty-five years rather than five or six. Over that twenty-five-year period, the [00:07:45] S&P five hundred with dividends had returned about eight percent a year. The [00:07:50] long-term average for the S&P five hundred is over ten percent a year. So using the [00:07:55] exact same logic, you could say we're behind average, so now we must be due for higher [00:08:00] returns to catch up.

Because for the last twenty-five years, we've made eight, the long-term returns are [00:08:05] actually twenty-five percent higher, going at ten percent a year, right? So, so, [00:08:10] so that would be the wrong conclusion, and it's no different than saying based on the recent [00:08:15] five or six years' returns, the market has to go lower.

Same kind of thing. [00:08:20] Extrapolating up or down leads to the same mistake. It creates a false sense that we, we [00:08:25] can predict what's about to come next, and that belief leads to decisions that can interrupt, [00:08:30] and often do interrupt, the investor's compounding. The direction of the extrapolation is [00:08:35] not what's important, but the damage that it causes is what is.

Context also [00:08:40] matters. In that twenty-five-year period, the market declined about fifty percent twice, [00:08:45] from two thousand to 'oh two and from two thousand seven to 'oh nine. That was actually a sixty percent decline. [00:08:50] Events that historically happen once or twice in a century happen [00:08:55] twice in less than a single decade.

So depending upon where you start the clock, you'll [00:09:00] see a different pattern. The pattern is real, but it doesn't make the future predictable. [00:09:05] Here's the real danger. Extrapolation drives decisions, and those decisions can [00:09:10] interrupt compounding. After strong patterns, people often pull back. After disappointing [00:09:15] patterns, people Try to reposition.

In both cases, they're [00:09:20] acting on a belief that the past tells them what the future will be. Compounding doesn't require [00:09:25] prediction. It requires discipline in the face of uncertainty to stay on the long-term plan. I'm [00:09:30] gonna share a little extrapolation story that I always felt was, uh, instructive. In, in the late [00:09:35] 1800s, there was something going on in cities like New York and London.

It was a [00:09:40] growing problem and actually considered to be an existential crisis, and it was called [00:09:45] the horse manure crisis of the 1890s. Everything depended on horses. Experts looked [00:09:50] at that pattern. The-- more horses every year, horses for taxi cabs, horses [00:09:55] for buses, horses for, uh, farming, et cetera. You name it, there were horses doing, [00:10:00] uh, doing most of the heavy duty tasks.

And they extrapolated the pattern forward. They [00:10:05] concluded cities would be buried in horse manure in short order. And it [00:10:10] wasn't a fringe worry. Newspapers ran serious pieces on it. You can Google it. City [00:10:15] planners lost sleep over it. It was treated as a near certainty, not a [00:10:20] possibility. There was even one of the first international conferences called in New York [00:10:25] to address these issues.

It was shut down in a matter of days because the entire [00:10:30] consortium who was debating it deemed that the problem was unsolvable. They [00:10:35] saw the pattern correctly. They were just wrong to assume it would continue [00:10:40] indefinitely. What they didn't see coming was the automobile, Ford and [00:10:45] Daimler. And just like that, the problem disappeared.

The lesson is not that those experts were [00:10:50] foolish. They were looking at the best data available to them. The lesson is that no amount of [00:10:55] past data can see around the corner. It doesn't know exists. [00:11:00] Extrapolation almost left them buried in something that we'd frankly all rather avoid. So what's [00:11:05] the takeaway?

The next time you hear returns have been high, so they must be lower going forward, or [00:11:10] returns have been low, so now's a good time to increase my exposure to equities 'cause they're gonna be higher, [00:11:15] just pause and ask yourself a few things. Am I observing a pattern or am I [00:11:20] projecting it forward? What timeframe is being emphasized and why?

Am I looking at the, the [00:11:25] appropriate denominator in the equation? Am I turning a long-term concept into a [00:11:30] short-term prediction? Would I reach the same conclusion if I started the clock five [00:11:35] years earlier or five years later? Because the danger isn't in seeing patterns. [00:11:40] It's believing that they can tell us what comes next.

At Fusion, we [00:11:45] actually believe strongly in using history as a guide, but in the right way. [00:11:50] History teaches us the nature of markets, that declines happen, volatility is normal, and [00:11:55] that long-term ownership has been rewarded over time. What history doesn't give us is [00:12:00] a reliable way to predict the next six, twelve, or even twenty-four months, [00:12:05] especially when we cherry-pick a timeframe to support a narrative, because that's the real [00:12:10] trap.

Conflating a pattern with predictive value. And it's that confusion [00:12:15] more than anything else that leads to decisions that interrupt compounding. [00:12:20] Fix It Friday generally is not about calling the next move. It's about fixing the thinking that [00:12:25] gets in the way of staying invested long enough for compounding to actually work.

[00:12:30] Thanks again for tuning in. Uh, you can catch us on [00:12:35] crazywealthypodcast.com, fusionfamilywealth.com, and all your favorite podcast venues. Until [00:12:40] next time, have a great weekend.

Voiceover: Thank you for [00:12:45] tuning in to another episode of the Crazy Wealthy Podcast. For more [00:12:50] insights, resources, and to sign up for our newsletter, visit [00:12:55] crazywealthypodcast.com. Until then, stay crazy wealthy[00:13:00]

Disclaimer: The previous podcast by [00:13:05] Fusion Family Wealth LLC, Fusion, was intended for general information purposes only. No portion of the podcast serves as the receipt of, or [00:13:10] as a substitute for, personalized investment advice from Fusion or any other investment professional of your choosing. Different types of investments involve [00:13:15] varying degrees of risk, and it should not be assumed that future performance of any specific investment or investment strategy or any non-investment related or planning [00:13:20] services, discussion, or content will be profitable, be suitable for your portfolio or individual situation.

Neither Fusion's investment advisor registration [00:13:25] status, nor any amount of prior experience or success, should be construed that a certain level of results or satisfaction will be achieved if Fusion is [00:13:30] engaged, or continues to be engaged, to provide investment advisory services. Fusion is neither a law firm nor accounting firm, and no portion of its services should be construed [00:13:35] as legal or accounting advice.

No portion of the video content should be construed by a client or prospective client as a guarantee that he or she will experience [00:13:40] a certain level of results if Fusion is engaged, or continues to be engaged, to provide investment advisory services. A copy of Fusion's current [00:13:45] written disclosure brochure discussing our advisory services and fees is available upon request or at [00:13:50] www.fusionfamilywealth.com.