Stock market forecasts are fine and good

Making fun of market predictions that end up being wrong can be fun. But being right is not the point.

Hello and welcome to I’m Late to This, a newsletter about things I haven’t stopped thinking about yet.

For those of you who recently signed up looking for a take about why I’m so bullish on Stripe or my argument that TikTok and Robinhood are on a collision course, I’m sorry.

Those takes do not exist and never will. At least not here.

The former view, of course, is the Substack consensus.

The latter either has or will be written, probably on a Substack near you.

So but for those of you who are new here, welcome.

We tend to publish most Sundays and write about things that are directionally about finance or markets, but not in a newsy way. Except when the letter is about current events.

Again, no promises, but, as happened last month, if I find myself just not that into whatever it is I’m sending out I won’t send anything. All of our inboxes are increasingly cluttered by Substacks these days and there’s an implicit agreement between author and subscriber that care has been taken with this piece of writing.

And while structure unlocks creativity and weekly has been a good cadence for this project, my own goal of publishing this newsletter isn’t going to supersede the respect I try to have for the inboxes of more than 1,900 people who are busy and, presumably, already over-emailed.

If you found this letter through Twitter or other channels and you haven’t done so yet, sign up below so you never miss an issue.

And if you like what you read, please make sure to share this letter with anyone you think may be interested.

I’m Late to This is free to read and plans to remain that way indefinitely.

And now, stock market predictions.

---

In the investment industry it’s year-ahead forecast season. 

Which means it is also “year-ahead forecasts don’t make any sense” season. 

And, look, I get it. 

There is nothing that anyone who works on Wall Street or anywhere else knows about what is going to happen next year that is worth sharing in any sort of definitive way.

But year-ahead forecasts are not definitive. It is only the critique of them that is.

And to suggest that sharing a view on what the next year might have in store for a company, the economy, or financial markets is dumb because no one is omniscient constructs a straw man. Last week, we referenced one of my favorite frames which comes from Chetan Puttagunta, who says his job as an investor is to the see the present clearly. A clean outline of what is happening now and why will naturally lead one to offer a view on how things will evolve. That view need not be closely held. But in saying that there’s basically no value in guessing at what the future holds you end up discounting the present, too. And things devolve from there.

Let’s take the most basic forecast out there in the investment industry, which is where the S&P 500 will trade in one year’s time. 

Every year, dozens of strategists at a variety of firms start thinking about this question some time after Labor Day and in late November or early December they roll out these house views to their clients. Here’s where a bunch of firms stand so far. Surprise! Everyone thinks stocks will go up next year for the reason we’ve covered previously.

These notes often span 50 pages or more. They are like small books, often summarizing not only the view of the equity strategy team but what every other major group at the firm — economics, industry groups, fixed income, commodities, and so on — is also thinking for the year ahead. 

If you’re a client on the receiving end of these reports, please let me know how useful you find them. I am assuming that “somewhat” is at least the median answer because anything below this threshold, you’d imagine, would have most research shops re-thinking the role these reports play in their workflow. 

Matt Levine has written extensively about just what it is sell side research is good for. In a March 2017 edition of Money Stuff, Levine outlined the model as follows:

What do sell-side equity research analysts do? There are three main theories:

1. They discover which stocks will go up, and then write research reports telling investing clients to buy them.

2. They lie about which stocks will go up in order to drum up investment banking business for their employers.

3. They are in an essentially client-service business of helping their investor clients by providing background research, deep-dive analysis, financial modeling, trade ideas, management access and, yes, sometimes Buy/Sell recommendations.

Theory 1 seems naïve: It's the investors' job to pick which stocks to buy, and it would be strange for them to outsource that job to research analysts with no skin in the game. Theory 2 is the knee-jerk cynical opposite of Theory 1, but it also doesn't seem to fit the facts. Theory 3 is, it seems to me, the realistic theory, describing a business that customers might actually want.

And so I know that readers of this newsletter signed up, presumably, for my take on things but in general I tend to outsource as much as possible. That was more or less the point of last week’s letter. 

Levine Theories 1 and 2 outline a truth wherein Wall Street research exists in a vacuum as either a conspiracy or a prophecy. And given how few things in the world are either, these seem unlikely answers. Even for an industry that has done as little introspection as Marc Rubenstein argues the sell side has managed to do.

So we proceed with Levine Theory 3, which says that research serves as a thing that exists for reasons that make sense in the context of a variety of investment processes rather than in a vacuum. Again, it is the “client-service business of helping their investor clients by providing background research, deep-dive analysis, financial modeling, trade ideas, management access and, yes, sometimes Buy/Sell recommendations.” 

To stylize this model another way, sell side research is like the sheet of answers you get along with a set of problems for a math class or something. The point of doing the problem set is not to get the right answer but to see if you’re learning the right process that will uncover the right answer under a variety of scenarios. Otherwise there’d be no reason to give students the answers. 

No one receiving a year ahead outlook from, say, Bank of America, doesn’t also have a rough idea of where they think the market is headed next year. In other words, this client has also done their own work on the same problem set and is comparing notes with BofA’s answer sheet. BofA, for what it’s worth, thinks the S&P 500 will trade up only modestly next year, rising to 3,800 from Friday’s closing level of 3,699. But no one cracks open this report and says, “My God, the S&P 500 is going to 3,800 next year!”

We’re not going to go through all the nitty gritty of why BofA is relatively conservative — through last week, the average forecast among sell side strategists was around 3,900 for the S&P 500 next year, though this average didn’t include at least one upgraded price target of 4,200 that came out last week — but the very basic outline of their view is that a lot of optimism regarding what 2021 could mean for stocks and the economy has already been priced in. No, there will not be a several thousand word digression on the epistemological underpinnings of this financial cliché. 

So but the way that Savita Subramanian or any other strategist arrives at this number is simple math: year-end price targets are the product of corporate earnings times the market multiple. In other words, earnings times how much investors will pay for one dollar of those earnings. BofA arrives at 3,800 by multiplying $165 in earnings for the S&P 500 by 23 (and some decimals). Again, everyone receiving this report is going to have their own view on both the earnings and the multiple. BofA’s answers merely offer another lens through which to check your own work and assumptions. 

Which gets to another salient part of why a critique against these reports falls flat — if year-ahead forecasts were as useless as the critics deem them to be they wouldn’t exist. 

In financial commentary there is a strain of criticism that hinges on arguments about the irrationality of some action taken by an investor — or some dynamic happening in the market — which suggests no one else had thought of the critique before taking the criticized action.

Like, if you buy debt issued by some country at a negative effective yield it seems irrational as a position. “Negative yields make no sense,” someone might say. 

But the answer to a question about why this nonsensical thing exists cannot be that everyone is choosing to be irrational. Otherwise there’d be no way for the negative yielding issue to come to market and successfully fill an order book. And yet that is exactly what happened. Views from the outside that someone else is being crazy, then, are wrong definitionally.[1]

That there is a defense of a financial industry habit which has gone on at this length does, on some level, disqualify the practice. Either something you do makes sense or it doesn’t. No one who sits outside the industry, even in the related-but-still-adjacent position that I do, should have to write a thousand words or more explaining why a basic practice is still a practice if this practice is any good. Sports reporters don’t defend drills at this length.

But the dissemination of these forecasts from Wall Street often comes through a financial media which doesn’t have their own work to show. And so whereas the Levine Theory 3 model of sell side research leads us away from seeing these reports in a vacuum, the context of all media is essentially a void. And while it’s a bit textualist of me to argue that no media story can exist within any larger contexts, the disconnect between the reader’s interest and expertise in a story and the reporter’s interest and expertise in a story can almost never be bridged. Unless you’re Matt Levine. Which is the reason that media criticism — like almost all forms of criticism these days — tends to focus on the author of the story rather than the story itself.[2]

So it then becomes a sensible argument from the cheap(er) seats to say that year-ahead forecasts are bad because they are often wrong. And especially so if the forecasts are presented as a kind of unimpeachable truth-telling exercise from some person or organization that cannot, in fact, see the future. But, again, that is not what this genre actually is.

The reductio response to this critique, of course, is that if your view is “no one knows anything so why should anyone have a forecast” you indeed have a forecast. The forecast is “no forecast” but that is an expectation. And not only an expectation but essentially a promise that when whatever happens happens you won’t be surprised. In Pulp Fiction, Uma Thurman says she can’t promise not to get offended by John Travolta’s question because he hasn’t asked it yet. The argument that “no one knows anything so why bother” says the opposite — I will never be offended or surprised by something that happens because anything could. I guess these people think the pandemic is no big deal.

And so here we see the value in thinking through why something exists.

Not because it teaches us much about the thing itself — though it does — but because it exposes reactions to the thing (in this case year-ahead price targets for the S&P 500) as offering different versions of the truth that can be even more potent.

Because in doubting that anyone can know the future it is argued that since the future itself is certain but the details are not I will worry about neither. 

Which is a claim much bolder than estimating where stocks might trade in a year.

---

1: Right, sorry, this only happens because central banks are manipulating formerly free markets. Silly me.

2: This is also the reason that subject area expert Substack-like publications have such a greenfield opportunity.