What I Think About The Goldman Prediction

The internet is abuzz with news of a recent prediction by Goldman Sachs that the S&P 500 could return 3% over the next decade.

I watched the Goldman US Equity Strategist David Kostin on Bloomberg explain the call.

One of the things the thesis hinges on is valuation. Currently, the S&P 500 has a p/e of 29.8, which is high but not high enough to justify the call. It’s funny but Warren Buffet once made a similar prediction about markets after someone asked him the following question: “The price-to-earnings ratio of the S&P 500 is significantly higher than the historical average. What benchmark should an investor use in purchasing this index?” Buffet didn’t answer this question but Charlie Munger said that while in the past he didn’t think indices could be so high priced that future returns would be low, he had changed his mind after seeing what happened in Japan. Buffet remarked that the S&P 500 would return 3-4% in such a scenario. In Japan, when the Nikkei topped in 1990, p/e was at 40. The question is, therefore, how high can p/e go before we have the S&P 500’s price be too high?

Various people have also commented that the S&P 500 only ever does 3% after a terrible recession i.e. the GFC, the dotcom crash, the great depression, and the oil crisis in the 70s. Even the covid crash was not big enough to lead to 3% returns. The funny thing is that during recessions, earnings go down and then p/e goes up. So valuations go higher in a recession. In May 2009, the S&P 500’s p/e was 123x, and this wasn’t a sign of a bubble. The historical average is 20x and even though we are above average, it’s not high enough to signify a bubble.

The thesis also hinges on concentration. Since the top 10 companies in the S&P 500 make up about 32% of the index, as this concentration subsides, then returns will be low. This is something that has been on many analyst’s minds. I recently read an interesting thought experiment on what happens when liquidity absorbed by the mag7 reverses.

I don’t think concentration matters either. Index investing is expected to be an outperforming strategy for the average investor because he or she doesn’t have the skills to pick winners—few people do anyway. Sometimes the winners are concentrated, and sometimes not, and you can’t tell beforehand. Also, the S&P 500 constantly adds and removes companies so that in the next 10 years, a third of the companies will be new companies. Right now it seems as if mag7 companies are the big winners in AI, but in 10 years it could be a different story in the same way that companies like Facebook (now meta) popped out of nowhere. You buy the index fund to gain exposure to the winners, even if they haven’t yet been born. The number of winners is irrelevant because the index is an average, but I would like to prove that assertion quantitatively. That is why I started to look into how to quantify what’s happening inside an index (i.e. breadth, momentum, etc.) but to continue that line of research I need higher-quality data which I don’t have1.

Predictions are hard and I’m not trying to rip into Goldman. What I take from this instead, especially the reaction to the prediction, is that people are starting to get really nervous about the SNP’s gains. I think that will translate to sharp declines in any events that signal the start of a recession, and those in the know will be buying exit doors and diversifying. The VIX is already quite elevated; currently trading at 19 which is a hair away from the all-important 20 figure which is the line in the sand often used to tell if people are freaking out. Considering that gold is already quite high, rates are coming down, and real estate is expensive right now, I’d say commodities would be the best choice for diversification. I’d be looking out for increased commercial interest in the commodity COT reports to signal the start of a trend higher in coms (NOT INVESTMENT ADVICE).