Bet Sizing...

Interested to hear others' thoughts on this
From Bloomberg: 

A lot has been said about Elm Wealth’s crystal ball challenge, where players trade the S&P 500 and 30-year Treasuries with leverage based on the Wall Street Journal’s front page. The idea is to test whether perfect foresight would actually yield stellar trading outcomes.

The answer among online players and 118 “young adults trained in finance” was no, partly because they got the market reaction wrong, but also partly because of poor trade sizing, or use of leverage. Five “seasoned and successful macro traders” did far better.

The game is a simplified exercise, of course, but I thought it was a neat illustration about a few themes you hear often in the investment industry.

While playing the game, I often chose to skip the trade because it felt impossible to guess the market reaction not knowing what was already priced in. (You could sort of extrapolate this from the framing of the articles, but not entirely.) This seemed especially true for the equity market, which the game’s designers noted were harder to get right compared to bonds for most players. (Most of the relevant headlines were about the economy, which struck me as having a more direct impact on long-end bonds than the S&P 500, which would be subject to many other forces.) This underscores why there’s so much interest in positioning and flows data.

Of course, in this case the five pros didn’t have that either. But Elm observed that they bet big on high-conviction days and skipped a third of the opportunities.

The other takeaway is that making the right bet and trade sizing are different skills. You hear this more often in the context of stock-picking. It’s part of the idea behind so-called “alpha capture,” which puts portfolios together based on (typically) the best ideas from sell-side analysts or fund managers. There’s been plenty of quant research on optimal portfolio construction, which is often now done by software.

The five mysterious pros were clearly also better at trade sizing, but some of this in the real world is also down to the constraints of your institutional set-up, like how big you’re allowed to bet on your highest-conviction ideas. You could have great ideas and insights and still end up with an underperforming portfolio.