Wall Street may be full of quants and algos now, but at the heart of investing and trading remains a human mind looking out over a landscape and assessing all the components and their interactions.

George Soros (set aside what you might read in right-wing propaganda for a moment) was famously an intuitive investor. His son once described how an onset of back spasms would cause George to turn cautious on risky positions — it was his risk meter.

Soros had a pithy heuristic: “invest, then investigate.“ 

He would buy on a hunch, and then start the process of research into the investment. This defies 100% of the advice you will get from your “wealth advisor”.  Of course we are supposed to load up on data and consider all the angles before risking one dollar. 

There’s only one thing missing from their advice: skin in the game. 

Soros knew from experience that there’s a world of difference between looking at risk and being at risk. Between being inside the experiment and observing it, between gaming and actually being shot at. Skin in the game is why military trainers fire live rounds over their recruits who are crawling along the ground with full packs.

Soros was not saying ignore data, he was saying put yourself in the best position to value the data by being exposed to the cost of misunderstanding it. 

Similarly, Nicholas Nassim Taleb, who writes about probability and sometimes does complicated equations to illustrate it, tells his readers that he did not start with the equations.

He started in options trading. With his own money at risk – skin in the game. He writes:

When there was risk on the line, suddenly a second brain in me manifested itself, and the probabilities of intricate sequences became suddenly effortless to analyze and map. When there is fire, you will run faster than in any competition. When you ski downhill some movements become effortless. Then I became dumb again when there was no real action. Furthermore, as traders, the mathematics we used fit our problem like a glove, unlike academics with a theory looking for some application—in some cases we had to invent models out of thin air and could not afford the wrong equations. Applying math to practical problems was another business altogether; it meant a deep understanding of the problem before writing the equations.

All this helps us understand why prediction markets are often more accurate than mere predictions or opinion polls. Kalshi’s prediction markets have been showing, repeatedly, more accuracy in forecasting market outcomes and event outcomes than opinion polls, journalists and media commentators. 

The most recent round of state level elections proved the point, which was not lost on many observers. 

Why is this? “Expert” public predictions without cost can include one or all of these biases: marketing, vanity, resentment, political preference, wishful thinking, attention-seeking, planned disinformation and manipulation. You can run the same review for the reliability of opinion polls taken by landline telephone during prime-time broadcast TV.

Skin in the game predictions include this: the need to be right.  

Little of what presents itself to your eyeballs these days is the product of a curious journalist wandering through the world in search of truth. The media doesn’t work like that anymore. Barring breaking, hard-news events – fires, murders, invasions – a high proportion of what appears in the media has been presented to the media by an interested party.

Ben Hunt, who runs an investment fund and the cultural inquiry website Epsilon Theory, has a useful guide to dealing with this: always ask yourself “why am I reading this now?”  He argues that planned narrative management has become so sophisticated, and so pervasively deployed, that we need to question the purpose of everything in the media, even if it may be prima facie true.

In such a world, predictions derived from event markets – the choices of traders with skin in the game – seems to be some of the cleanest information around.