How to forecast the stock market?

Don't bother

The ups and downs of the US stock market make for daily business headlines. Sometimes, most often on big down days, the headlines are on the front page. Newspaper editors find crashes much more newsworthy than booms, revealing their predilection for bad news, which in turn explains why news junkies are very poor dinner party guests.

The biggest equity market in the world is fascinating (for those who need to get out more) for its own sake and because, over the short term at least, it is a big driver of all stock markets. If you can guess which way it is going you can make an awful lot of money. Thats a big if.

Every neophyte stock market guru is always enchanted by the tools available to make forecasts. And convinced of the unique way they deploy their wisdom and massive intellect to the task. The tools can be very simple - drawing straight lines on charts, for instance. Or as complex as you like, deploying stochastic calculus and the latest non-linear regression techniques. Very smart people try their hand and fancy their chances. They all end up, sooner or later, doing something else. Hopefully something that adds a little value to humankind.

One of the best finance/economics blogs out there is the FT’s ‘Free Lunch’. Sadly, it’s usually behind a paywall. On Sundays it is written by Tej Parikh. Tej, normally brilliant, usually chooses his subject well and offers both insight and something interesting, if not new. His latest column forecasts a significant decline in the the US equity market between now and the end of the year.

He begins by saying:

End-year stock market forecasts are ultimately underpinned by investors’ annual economic outlook and their assessment of structural drivers, such as artificial intelligence and US exceptionalism.

If he had substituted ‘sometimes’ for ‘ultimately’ there would be little to disagree with this opening sentence. Other than the fact that, written either way, it assumes we know what (a) is currently driving the stock market and (b) we know what those drivers will be at the end of the year. There is explicit date fetishism involved: what is with the obsession with December 31st?

He might be right about the structural drivers. AI for instance. It seems a reasonable thing to say, given the way AI news has indeed been a big driver in recent times. But, not so long ago, it wasn’t. And it may not be for much longer. And if it is a big driver, that’s not the end of the story: even when we can identify a factor that will affect the market, what will that factor be doing? Up or down? Big or small? On December 31st?

Naturally, Tej talks about ‘economic fundamentals’. He’s pessimistic about the market because he’s gloomy about the economic outlook. We might be tempted to point out that everybody, apart from MAGA fundamentalists (a dwindling band according to the polls) is pessimistic about the economy. Who is going to fall off their chair with surprise if there is a Trump-recession? Try finding a correlation in historic data between the economy and the one-year ahead performance of the stock market. If there is anything in the data (there probably isn’t) it suggests that markets, as we might expect, get there a lot earlier than economists and market strategists.

Tej cites strategists who argue equities are not priced even for a mild slowdown. That’s derived from a ‘simple regression model’. Plot a lot of dots on graph paper and start playing the children’s dot-to-dot game.

Tej asserts:

Expectations for corporate earnings this year remain too high.

That suggests Tej is new to game of looking at analyst profit forecasts. If he was an old hand he would know that earnings forecasts at the start of any year are almost always too high. It’s in the data for nearly every year going back to the dawn of time. Or at least when equity analysis started drawing in graduates who would otherwise have been employed doing something societally useful.

More importantly, Tej doesn’t realise that 100% of professional investors don’t believe analyst forecasts.

There is some analysis of what will happen if the Trump tariffs have to be absorbed in profit margins rather than passed on to consumers. Yes, profits will fall. Usually that’s not good news for share prices. Again, we might suspect that one or two investors may have figured this out already. Moreover, proper investors will ask if the tariff-induced fall in profits is likely to be permanent or temporary. If the latter, we can expect short-term volatility that will help traders but will leave proper investors unmoved. And only Peter Navarro believes tariffs will be both permanent and that the world won’t figure out to live with them.

Some investors might think that the slowing economy will elicit a big fall in interest rtes and bond yields. Something that could cancel out the lower profits thing.

Proper investors know that strategists have been wittering on about US profit margins for decades. That’s because those margins are consistently above their long term averages. And analysts are very fond of mean reversion: they look at something above its average and forecast it will revert to that average. Without ever pointing out that even if the data is consistent with mean reversion, that can happen in a number of ways, may take decades to happen and, even if it does, may not last very long.

Tej talks a lot about Price-to-Earnings ratios that are ‘too high’. There really is a lot of mean-reversion going on here. I really don’t know any professional investor who thinks a P/E ratio tells you anything - about an individual company or a whole market. Don’t just take my word for it, here’s Aswath Damodaran, the leading finance professor who really has written all this to know about this stuff:

Market indicators such as price-to-earnings (P/E) ratio, correlation and regression do not always work…I have not found these market indicators working as an investment strategy

We are then told that Bank of America’s Fund Manager survey revealed in its most recent edition that asset managers ‘slashed their US equity holdings by the most on record’. I guess they must have sold their stocks to non asset managers.

Somebody needs to tell Tej that most asset managers either assign filling out the survey to the most junior member of the department (who probably has no idea what the firm as a whole is doing) or wouldn’t dream of publicly revealing details of their most recent investing activity. In any event, Tej should ask himself this: has the BofA survey ever been a successful consistent predictor of anything?

Tej concludes with a bit of arm waving, something that all seasoned equity strategists are masters of.

The constant churn of policy announcements, exemptions, postponements and denials mean investors re-price each day where they consider risk to be relative to the day before. This then shifts the goalposts for judging growth and profitability forecasts…..for all the noise, however, the market still seems positioned for a hopeful outcome. Stocks are not even priced right now for a mild downturn

This is Hayekian ‘pretence of knowledge stuff’. Or maybe just word salad.

The arm waving ends with a firm forecast, something no seasoned equity market forecaster would ever do: the S&P will end the year below 5000. It’s currently 5525. So, a fall, but not a crash that will make headlines, at least not front page ones.

Tej might be right. If he is, then I will be the first to congratulate him. But, I would assert, he needs a lot of things to go his way and, even if the S&P is a lot lower at the end of the year, it will not be for the reasons that Tej suggests.

There is no known technique that allows for consistent, accurate forecasting of the stock market. The final thing that Tej needs to know is that market strategists are employed to get their names in the papers, to get advertising for their firm. No professional investor ever listens to a word they say.