Can collective intelligence be used to predict the future?

Introducing the original idea behind the crowdsourced financial forecasting Solex.AI Platform by our founder, Alex Merkulov.

This is a meditation on collective intelligence based on my experiences in the investing industry where I have spent the past 14 years observing highly paid professionals using an incredible amount of data, technology and the latest know how to predict the near-term future on a daily basis.

It is a fascinating feat to watch. This day and age, the sheer amount of information that is available to investors (and to you, if you pay enough attention) is astounding. It never pauses, it is increasingly global in nature and complex.

A price of a single financial instrument can be impacted by multiple seemingly unrelated factors. One day the main driver of change is someone’s interview, another — the news of a new central bank policy half the world away and the very next one — a new virus strain. Market sentiment fluctuates like the attention span of a teenage TikToker in search of a new trend. Some trends seem significant but stay with us only for a few days, some seem inconsequential but persist for much longer.

Trying to absorb and process all of this is very much like trying to drink water from a fire hose at full blast. There are data providers that can tell you what every single ship of earth is carrying and pinpoint its exact location; algorithms that scan social media pictures to know how often a certain t-shirt from a particular brand is worn; journalists reporting on facial expressions of politicians walking out of an important meeting. If you want to invest consistently well, prepare to be married to the news flow and it is a partner that does not wait.

One would think that with all this, professional investors cannot be wrong. Wrong. In the investment world, if you can get 55% of your bets right (not even the level — just the direction — up or down) you are considered good. 60%, or 6 out of 10 calls, on average, is considered an outstanding result, enough to be entrusted with billions of dollars.

Something to think about, huh? 6 out of 10 and billions.

Aside from the hit ratio, consistency is valued. Few will invest with someone who has a brilliant result once in a blue moon and mediocre performance otherwise. Having consistency means being predictable. Predictably good to be exact, but no one stays good forever.

Being a good investor or a financial advisor is not dissimilar to being a good athlete — you must consistently be on the edge. Tales of legendary investors becoming too comfortable and blowing their careers up with a few bad calls are abound. Humans age, get divorced, skipped for a promotion and angry. In short, life happens and all of it can impact the quality of their decisions, no matter how good they used to be.

For these and many other reasons, picking who you trust, including yourself, is a serious matter and one that cannot stay static.

If you have enough time and capital, you can hire multiple advisors and generally do ok with respect to the issues I have outlined above. Most of us do not, however. To me, this is a challenge and I believe an opportunity.

So, predicting the future is hard (duh) and no single person is likely to do well over a long time. So what option do we have?

I believe the answer lies in collective intelligence.

What if you do not have to rely on expertise of one (or even the best few), but instead tune in to thousands of market watchers connected by an algorithm that can show you what they all think on average, right now? What if that algorithm can tilt towards predictions from users who are actually good at calling the market statistically, not because they happen to have a certain degree, a family connection in the industry or live in a certain area?

Prediction markets have shown to be more accurate than polls in politics. Collective intelligence has been used to locate USS Scorpion, a sunken nuclear submarine, that vanished without a trace in the open ocean with astonishing accuracy. The CIA has conducted a number of experiments in late 2000’s that showed that diverse crowds are better at predicting terrorist acts than a small group of highly qualified experts. It is promising in many other fields.

Markets themselves are a collective intelligence. The price of everything you buy is largely determined by our combined best guess as to what that is worth. Some very large markets have a similar predictive mechanism in a form of futures or forwards, but they are few and one needs deep pockets to play there.

It is my firm opinion that we can do better by opening a possibility for everyone to get involved, let collective intelligence drive the consensus and let statistics be the judge of the accuracy of our forecasts. More markets can be covered and more people can take an active part. Everyone can benefit from accessing the data for free, in real time. If we can consistently demonstrate that collective predictions match or beat the expect opinion, we can and should financially reward those serving the collective good.

At the very least, if we can achieve or beat the 6 out of 10 hit ratio, those who charge for advice would have to outperform the collective us!