The Better Letter: Forecasting Follies
The Better Letter: Forecasting Follies
Confident 2020 predictions gone wrong...as ever
How do you measure a year?
In science, success is measured by predictive power. A valid scientific theory can be applied to predict real phenomena. I will apply that test to market predictions this week in my annual look at forecasting follies.
Spoiler Alert: They aren’t very scientific.
More broadly, this is the season of measuring the year just past. By nearly any measure, it was a lousy one.
So, even though 2021 has gotten off to a rocky start, Happy New Year! And good riddance to 2020.
If you like The Better Letter, please subscribe, share it, and forward it widely. It’s free, there are no ads, and I never sell or give away email addresses.
Thanks for visiting.
Forecasting Follies 2021
In early October 1997, it had been just over two months since Steve Jobs returned to the company he founded to take over as CEO from alleged turnaround artist Gil Amelio. During Amelio’s 500-day tenure, Apple suffered $1.6 billion in losses, essentially wiping out all the profits the company had earned since 1991.
Jobs was preparing to deliver the final bit of news from Amelio’s failed leadership: Apple would report a loss of $161 million, the largest quarterly loss in the company’s history. Other than report the loss, the only other thing he had to announce was a new marketing campaign.
It wasn’t a great time.
Understandably, other tech leaders were skeptical that Jobs could save the faltering company. During that time, billionaire computer maker Michael Dell was onstage at an industry event for a Q&A session with reporters and was asked what he would do with the struggling Apple.
“Shut it down and give the money back to shareholders,” was his reply.
Jobs responded with an email to Dell noting Dell’s lack of “class.” His real reply came a few days later at a company town hall.
We now know, indisputably, that Dell was wildly wrong. Through the end of 2020, investors in Apple stock earned returns of 81,500 percent (nearly 34% annualized) since Dell’s infamous comment (thanks, Jake, for the data calculation). In other words, a $10,000 investment contra Dell’s advice on the fall day in 1997 he gave it would be worth more than $8,500,000 today.
It’s fair to say that Dell’s “analysis” was so flawed it’s “not even wrong.”
In 2010, five U.S. equity fund managers were nominated to be the fund manager of the decade for their outstanding performance over the previous ten years. Over the ensuing decade, exactly zero of the five outperformed. The average subsequent underperformance of the five nominees was 5 percentage points per year. The “best” fund manager of the decade (through 2009), over the following decade (through 2019), underperformed by 8 percentage points per year.
ESPN polled its panel of basketball experts for their predictions about the upcoming 2019-2020 NBA season. These are people who were hired for their basketball knowledge and whose entire job is to be an expert about basketball. Most of them picked the Los Angeles Clippers. Some of them picked the Milwaukee Bucks. Two picked the Philadelphia 76ers. Not a single one of them thought the Los Angeles Lakers would win it all.
Guess who did? The ESPN prognostications were “just a bit outside.”
From time immemorial, people have sought to play God, even to be God. We’re terrible at it. As the great polymath Freeman Dyson explained, history is replete with those “who make confident predictions about the future and end up believing their predictions.”
Daniel Kahneman highlighted the “illusion of understanding” – our predisposition to concoct stories from the information we have on hand. “The core of the illusion is that we believe we understand the past, which implies that the future should also be knowable.”
However, history is driven by surprising events while forecasting is driven by predictable ones. Let the great Peter Bernstein explain more precisely (Peter L. Bernstein, “Forecasting: Fables, Failures, and Futures – Continued,” in Economics and Portfolio Strategy, November 15, 2002, p. 5).
“The very idea that a forecaster can spin a bunch of outcomes whose probabilities add up to 100 percent is a kind of hubris. Risk means that more things can happen than will happen, which in turn means that the scenarios we spin will never add up to 100 percent of the future possibilities except as a matter of luck. Like it or not, the unimaginable outcomes are the ones that make the biggest spread between expected asset returns and the actual result.”
As Jason Zweig put it, “The only true certainty is surprise.”
The coronavirus crisis took everyone by surprise, even though a global pandemic of this sort has been known to be highly likely as a general matter for quite some time. It’s always something. Still, a prominent scholar was certain COVID-19 deaths would be limited to 500 or so. I wish he had been right.
Hundreds of researchers attempted to predict how various children and families would turn out, using 15 years of data (the full research, published in 2020, is here). None was able to do so with meaningful accuracy.
In late October, Pat Robertson declared that he had heard from the Lord: “Without question, Trump is going to win the election.” He (or maybe the Lord, though I doubt it) was wrong, obviously, as were those sure about the market consequences of the election.
As the mathematician Alfred Cowles observed decades ago, people “want to believe that somebody really knows.”
Nobody does.
As John Kenneth Galbraith succinctly (and correctly) stated, “There are two kinds of forecasters: those who don’t know, and those who don’t know they don’t know.” Stock market forecasts prove its truth year after year.
2020 was no different.
In December 2019, the median consensus on Wall Street was that the S&P 500 would rise 2.7 percent in the 2020 calendar year. That target turned out to be way too low as the S&P 500 earned 18 percent for the year. That’s a forecasting error of more than 15 percentage points — more than five times the estimate of the market’s increase for the entire year.
That would have been bad enough, but when the stock market plummeted in February and March, forecasters wouldn’t leave bad enough alone. In April, the Bloomberg survey showed, forecasters “recalibrated” and predicted that the S&P 500 wouldn’t rise at all in 2020. Instead, they said the market would fall about 11 percent. However, the market had already begun climbing on March 23, the day the Fed intervened to stem panic. The “expert” strategists failed even to register the change in direction and their “recalibrated” predictions were off by nearly 30 percentage points.
Last January, Guggenheim Partners’ Scott Minerd told delegates at Davos that world markets are a “Ponzi scheme.” That was quite a claim, obviously, and suggested that he thinks we are in the final stages of the progression that the late Hyman Minsky diagnosed, in which increasing debt-fueled speculation leads ultimately to collapse. He called for “another Great Depression.” Nope.
On March 2, fixed income specialty firm Western Asset Management called a top in U.S. Treasuries. The benchmark 10-year U.S. Treasury note closed that day yielding 1.10 percent. Just one week later, 10-year notes had rallied to yield an astonishing 56 basis points. Nope.
Nuveen strategist Bob Doll expected 10-year note yields to exceed two percent in 2020, stocks and bonds to return less than five percent each, active managers to outperform, and President Trump to be reelected. GMO called for negative stock and bond returns. Nope times two.
One “analyst” advocated owning only hard assets and shifted all his focus to “survivalism and prepping.” And nope.
On Sunday, March 8, a prominent financial journalist began making market predictions as the threat of COVID-19 became abundantly clear. He expected the market to drop 25-30 percent over the following three to four months. When the S&P 500 plunged more than 7 percent the next morning, he revised his time frame to three to four hours and added, “Good luck trying to catch this falling knife.”
The following Monday, with the news growing even more ominous, he updated his prediction. The Dow was heading to 15,000. When the Dow dipped below 20,000, he made 10,000 his new target. The Dow dropped another 1,338 points the next day and the prognostication surely felt right.
But then, on March 23, the day the Fed intervened to stem panic, the “Otter defense” …
… which had seemed so unlikely, took hold, and stocks began to rise. Seemingly inexorably.
More than 4,000 Americans are dying each day from COVID-19 this week. Sometime after 10 am PT Wednesday, a mob of insurrectionists managed a successful large-scale breach of the U.S. Capitol for the first time since the War of 1812, while the S&P 500 hit a new all-time high above 3,780. Smith & Wesson stock rallied more than 17 percent. Yesterday (Thursday), the S&P, the Dow, and the Nasdaq all closed at new all-time highs.
I wonder what might have happened had the news been good.
If you had been in cash at the market bottom in March and stayed there, you would have missed out on a 68 percent return on the S&P 500 through year-end. If, instead, you had bet on further market collapse and bought the ProShares UltraPro Short S&P500 ETF, you would have lost -80 percent through year-end (h/t to Jason Zweig’s indispensable newsletter for the $SPXU idea and Jake for the return calculations).
Granted, the rally followed a steep drawdown, but it has far exceeded the extent of that drawdown. Despite a global pandemic, the worst economic downturn since the Great Depression, and a bitterly contested election, the S&P 500 made 33 all-time highs in 2020, the 33rd on New Year’s Eve, the final trading day of the year.
In 1988, Warren Buffett wrote the following in his annual letter to shareholders.
“As regular readers of this report know, our new commitments are not based on a judgment about short-term prospects for the stock market. Rather, they reflect an opinion about long-term business prospects for specific companies. We do not have, never have had, and never will have an opinion about where the stock market, interest rates, or business activity will be a year from now.”
The great Benjamin Graham made a similar admission in a 1963 speech (emphasis in original).
“I would like to point out that the last time I made any stock-market predictions was in the year 1914, when my firm judged me qualified to write their daily market letter, based on the fact that I had one month’s experience in Wall Street. Since then I have given up making predictions.”
As so often happens, Kahneman said it best.
“It's not that the pundits do badly. It's not that the television chains made a mistake. They didn't make a mistake. The world is incomprehensible. It's not the fault of the pundits. It's the fault of the world. It's just too complicated to predict. It's too complicated, and luck plays an enormously important role.”
As Kahneman emphasized, “Claims for correct intuitions in an unpredictable situation are self-delusional at best, sometimes worse.”
Since 2000, the median Wall Street analyst forecast has had the S&P 500 rising 9.5 percent per year, on average, which is close to the index’s historical average annual return. Over that time period, the annual increase of the S&P averaged 6 percent a year. This aggregate “miss” of 3.5 percentage points per year is bad, but it is even worse than it might appear.
Each December, since 2000, the median analyst forecast never called for a stock market decline over the course of the following calendar year. Not once. However, the market did fall in six separate years during that time. That’s roughly in line with the long-term stock market averages. From 1926 through 2020, the stock market fell in 26 percent of calendar years. All told the median Wall Street analyst forecast from 2000 through 2020 was off by an average of 12.9 percentage points, more than double the actual average annual performance of the stock market during that 20-year period.
As an important academic paper on forecasting showed, “In … economics we are faced with … a need for accurate forecasts, yet our ability to predict the future has been found wanting.”
So, when you come across a market forecast, “Pay no attention to the man behind the curtain.”
In fact, the history of market predictions is so bad, it’s tempting to use them as contra-indicators. When the consensus seems sure of a particular outcome, you might do the opposite.
You may be troubled by the vertiginous heights the market reached in 2020 despite a global pandemic and an economy that was largely shut down for significant periods. I am.
You may not be comfortable with current market valuations that have risen to nose-bleed heights. Me, too.
You might worry that markets are too dependent upon the Fed’s largesse and at risk going forward. That’s fair.
However, and given the forecasting records of every Wall Street “expert,” those concerns shouldn’t be pointed to as reasons to try to time the market or “go to cash.” One thing I can forecast with virtual certainty: There will be bad times and ugly losses ahead. I just don’t know when.
And neither does anybody else.
If you really could foresee the future, wouldn’t you be working at a hedge fund and making a fortune that way instead of making forecasts and doing media hits?
Because I believe in America and in the American economy – and the global economy, too – I remain invested in stocks and will stay invested in stocks. I could be wrong, of course, but the bias favors growth.
That’s not much of a forecast, but it’s the best I can do.
Totally Worth It
The consensus of expert forecasts for 2021, in a handy bingo card format, is available here.
If you’re relatively new to TBL, or merely want to read something interesting, have a look at these TBL missives from 2020.
Here is the best thing I read in 2020. Interesting defaults. 2020’s science superlatives. This is the most beautiful thing I saw last week (h/t Kenneth Scigulinsky). The sprightliest. The wisest, unless it was this. Why we make resolutions (and why they fail). RIP, Tom Perrotta.
Benediction
There is something special about human voices singing together. It is universal. Collaborative. Bonding. Healing. Joyous.
In church, I am most often stirred by the singing.
Freddie Mercury and Queen’s 21-minute Live Aid benefit set on July 13, 1985, is often called one of the great live performances of all time. Watch and listen to the huge and enraptured Wembley crowd sing-along, roaring back every call-and-response. It’s no wonder Live Aid raised $127 million for famine relief.
This week’s benediction is a virtual choir singing a 19th Century hymn that asks an appropriate question.
My life flows on in endless song; Above earth's lamentation, I hear the sweet, though far-off hymn That hails a new creation | Through all the tumult and the strife, I hear that music ringing It finds an echo in my soul How can I keep from singing?
Contact me via rpseawright [at] gmail [dot] com or on Twitter (@rpseawright). Don’t forget to subscribe and share.
Thanks for reading.
Issue 46 (January 8, 2021)