What would be the effects of (relatively) widespread precognition on the stock market?

I am working on a story in which around 1% of the world population has at least some semblance of a metahuman ability. Many metahuman abilities are advantageous in traditional employment, and I imagine that many of those who possess such abilities would put them to economic uses. For precognition users (henceforth abbreviated "precogs" in particular) there would be anywhere from several hundred to a few thousand of them. Some of them would probably be picked up by hedge funds, trading firms and other financial actors. I wonder what the effects of precognition on the stock (and other financial) market(s) would be.

Some Points of Note

The below is not necessarily accurate for all precogs in my setting, but it's the rules of thumbs that apply to most precogs in the verse and is probably sufficient for the purposes of this question.

• Precogs cannot anticipate their own decision (they can anticipate truly random processes e.g a quantum die) and can anticipate the world when they take a given course of action.
• Most precogs aren't clairvoyant and receive their sensory stream at some point in the future (so they cannot see beyond their death for example).
• If two precogs $$X$$ and $$Y$$ try to anticipate each other, only one of them would correctly anticipate the other. Let the probability that $$X$$ can anticipate $$Y$$ be $$X(Y)$$, and the probability that $$Y$$ can anticipate $$X$$ be $$Y(X), X(Y) + Y(X) = 1$$. This relation is transitive. $$X(Y) > Y(X), Y(Z) > Z(Y) \implies X(Z) > Z(X)$$ I think this can generalise to situations involving multiple precogs but I might be wrong. The above restriction applies when precogs are anticipating events impacted by other precogs even if not directly anticipating the actions of other precogs.
• In general, most precogs are short range (80% have a max range under a minute, 90% under an hour and 99% under a day).

Questions

1. What major effects would the employment of precognition by financial actors have on markets?
2. Would markets be exploitable to precogs?
• Are you using mutually exclusive precognition? AKA the Dune model wherein precogs cannot estimate each others' future actions?
– Ash
Jul 21, 2019 at 13:58
– Ash
Jul 21, 2019 at 14:23
• @TobiAlafin You've already accepted a solution. You generally want to wait at least a couple days before accepting a solution to give the community time to weigh in. In this particular case, I'm not sure the answer, but (as I said in comment), I'm pretty sure the currently accepted answer cannot be right. I'd like to see more discussion before the solution gets accepted.
– SRM
Jul 21, 2019 at 18:12
• Day trading is a zero sum game, so after a while of outsmarting each other, the net result would be minor. More profound result would be in emergency services, as knowing in advance about an accident may prevent it or at least buy a few crucial minutes. Jul 22, 2019 at 12:20
• Recommended reading, McCaffrey's To Ride Pegasus where these effects on wider society are discussed, including fields related to trading such as insurance.
– Nij
Jul 22, 2019 at 19:43

Within months, the Federal Government would ban them

All those that didn't get killed by Vegas, anyway. The best use for precog powers isn't the stock market, it's high-stakes gambling. Even someone with a fifteen second window would be a master gambler, constantly winning money at any casino of their choice. At which point the casino would get fed up and kick them out. Possibly after roughing them up. I have no doubt a black market ring of precogs would exist as a voice-in-the-ear for gamblers, but that would be illegal and just falls under the jurisdiction of 'massive profit made by illegally gambling'.

The fact is that these precogs would make excellent day traders for quick profit, but lacking the long term ability, they wouldn't be able to invest heavily into startup companies that turn out to be massive successes (i.e. Microsoft, Amazon, etc.) With that kind of power, they'd quickly start irritating everyone else on Wall Street by upstaging them. And these are the kinds of people you do not want to upstage. So the rich blokes on Wall Street would then run a campaign, explaining how the very existence of these people trading on Wall Street is ruining the system and preventing ordinary people from making a profit. Then laws get passed, forbidding precogs from exploiting the normal folk with their power.

The ones with actually power, your 1% of a 1% who have longer than a day would be employed by altogether different people which would make them more money. Major companies would have them on hand when making decisions, governments would employ them to see the results of special operations teams, and a few of them would probably start cults and make money that way.

• @TobiAlafin Major firms would hire the lesser ones. They'd try to hire the better ones, but there are far better opportunities out there than being a day trader for the top precogs. Until the government banned them. In which case they'd basically be courting the might of the federal government. The only way to not get caught is to not being doing well enough to serve as a statistical outlier - and in that case, what's the point? Jul 21, 2019 at 14:47
• They see their future sensory stream at time t+x, so they can't see remotely. They can access their own futures. They just can't see predictions of the form: (what would I choose in situation x (as such predictions might be paradoxical)). Jul 21, 2019 at 18:52
• @Ivella There's no constitutional provision for meta powers. (In all likelihood though, the case would go to the Supreme Court and become a huge political mess. But I'd bet that the public perception would be against metas at that point, given that all it takes is one idiot trying to be a hero to screw everything up, cough Marvel Civil War cough) Jul 22, 2019 at 17:03
• I could see them being employed as financial advisors but I concur that laws would be written to keep them away from the stock market as it would be akin to insider trading. How you'd police it though I've no idea given you can engage in trading from home. Jul 22, 2019 at 23:18
• Came here to say this, knowing what's going to happen to the markets ahead of time is called insider trading and it's already illegal
– Josh
Jul 23, 2019 at 12:39

A lot, followed by almost nothing. The first couple of minute or two range precog traders are going to have a big impact, until they start running over each other's trades, that would create some huge swings if they were high volume traders. Once a couple of dozen of them were all tripping each other up the total effect would start to cancel out any advantage they would have over normal people. The longer range precogs would probably keep their advantage longer but they're still going to get snowed out by other precogs doing things that look really odd at the time but turn out unexpectedly well.

• So a new precog wouldn't be able to easily loads of money on the market? Jul 21, 2019 at 15:10
• @TobiAlafin How much any precog can make in the market, over and above what a reasonably canny human can make would be directly proportional to the total number of precogs in the market, the more precogs there are the less likely an individual precog is to be able to outguess all the others.
– Ash
Jul 21, 2019 at 15:12

The efficient markets theory says that markets fully price in new information almost immediately. Applying this to precognition, information would become available sooner. But once available, it doesn't actually change the fundamentals of the market. It just discovers them quicker. So just from the perspective of trying to predict the market, precognition won't matter.

Now, from the perspective of trying to get a job in the market, that's a different story. Because precognition is better at detecting the fundamentals of the market than analysis, the precogs would take over all the financial jobs. This would drive out the non-precogs from those jobs.

Another issue is that certain kinds of investment would be more heavily impacted. For example, consider the following:

1. Where should I drill or mine? Precogs could pick the correct locations better than sampling (drilling holes to see what's in them).
2. Will this drug or treatment work? Testing new drugs or therapies takes more than ten years. A long range precog could simply predict the results. Even shorter range precogs could tell more immediate effects.
3. What will be the results of this experiment? If you have an expensive to run experiment, you could just ask a precog what would happen when you did the expensive step. Even a minute might be enough for some experiments.

A significant portion of investment activity is related to guessing what activities will be successful. If the precogs can eliminate the guessing, then it makes such investments less risky. And if they're less risky, then people will charge a lower premium to invest there.

So what will end up happening is that the precogs will control all the investment activity. Investments will operate more like banks. People will put money into their investment account for a clear promise of a specific return. The precogs will make sure that the investments perform well enough to support that return.

The days of people starting businesses in their garages will be over. If people could do this, the long range precogs would recognize their abilities quickly and send someone to offer them financing. If they refuse, then the precogs can find their optimal competitor and give that person financing, smothering the prospective business.

• Your opening paragraph cannot be true. New info becomes available from person X causes market to rise. Market adjusts. Y sees that the market is now going to fall and changes to buy. Market adjusts. The Efficient Markets here has to be as if new and contradictory info were continuously introduced. There's a calculus limit waiting to be computed somewhere in this question.
– SRM
Jul 21, 2019 at 18:11

The stock market is reactive

By looking at the future, the future is changing so every time a precog looks, the stock market will change

So when Mystical Mary looks to the future, she sees Apple going up and IBM going down so when Barry the Magnificent looks he sees people selling off their IBM and buying Apple so people sell their Apple to buy cheap IBM.

Eventually everyone is looking and the stock market goes crazy and nobody has any idea what will happen.

State Monopoly on Power

Every government is based on the idea of certain actions being exclusive to the state. You aren't allowed to print your own money, or build bombs in your basement, or own a tank. This is because those things would erode the power of the government to rule effectively. Your metahumans would most likely fall under the same category, at least partially. As such the government would have a vested interest in controlling how and when their abilities are put to use.

In the case of your precogs and the stock market, I can imagine the government going for a private-sector ban on using metahumans to influence the market. We actually already have a term for when someone uses knowledge about future decisions or information to make choices in the present. It is insider trading. Typically the "insider" has some direct tie to the company and that is how they get their information, but it isn't unreasonable to apply those same rules to precogs doing high-value day trading.

Contracted Cognition

You can decide exactly how scary the metahumans are, and therefore how big of a reaction their discovery generates. On one side of that scale is the "dangerous powers trump personal rights" argument. This would see all metahumans at least registered with the government. Most likely you would also have the most powerful/useful conscripted by the state for different jobs.

If you instead want things to be a little less dystopian you could go with the "metahumans are still human, just with special skillsets" form of governing. In that case you would still probably have registration, but maybe make it highly suggested instead of legally mandated. Also your metahumans would be offered lucrative government contracts that don't involve the barrel of a rifle...

In both instances what you end up with is the government having the largest share of precogs. They would be put to different tasks, but at least some of them would just be set to monitor the stock market and watch out for influence by rogue precogs (or supersmart human computers, or unscrupulous mindreaders, or anyone else who would impact their predictions).

Based on how you described multiple precogs affecting each other, if the government had enough of them then they should have a mathematically sound blockade around the stock market. Each precog notes their predictions for whichever time period they are strongest at, and then compares that to reality. If something doesn't line up then you have a pretty strong indicator that someone is trying to cheat. Since the government precogs are just making predictions and not trying to affect anything I am assuming that they would not interfere with each other.

Seedy Underbelly

The same basic thing as above would happen with less legitimate industries. If metahumans have to register with the government you would have a pretty good business with rogues selling their services on the black market. If metahumans have more rights then the same thing happens but with slightly more above board paperwork. Anti-precogs would find work in casinos and other gambling businesses, and would almost be a necessity. But you could also have metahuman bouncers or entertainment just as easily.

• Yeah, it's called insider trading, +1. But what would happen if everyone had access to all 'inside' information? I'm not exactly sure. The market would crash and make taking small businesses public, impossible? Which would doom ourselves to complete and utter 'corporate Americanism' of any and all corporations large enough to survive. So dystopia, in short order is what would happen. At least you wouldn't be able to 'short' investments anymore.... This would set economics back to before the first market in Amsterdam, c. 1602, because they wouldn't work anymore. Jul 22, 2019 at 23:41
• @Mazura If everybody had all of the information it would probably balance out to how things are now, except a little faster. But the OP specifically states that only some people have powers and only some of those powers can be used to cheat at the stocks. Since it isn't widespread it is an unfair advantage, just like normal insider trading. Jul 23, 2019 at 12:15

The very rare precogs that can see farthest into the future will be hired by insurance underwriters (< 0.001%) The farther you can see if a particular undertaking will need a payout the more valuable they’d be to that market

The shorter time frame precogs would not effect the practice of High Frequency Trading.

Short term day trading would nearly stop since everyone knows that there is a good likelihood that the person they are making the trade with is a precog and they are making the best possible decision for themselves, which means not good for you.

Long term, Value investing and dividend investing would become more dominant practices since speculative trades would be much more rational for the same reason as before.

I think anytime someone is investing on a hunch in a world with precogs they’d soon learn they can’t win, or loss the majority of the time. But markets based on capital growth and market innovation would have such long time horizons that precogs couldn’t pick winners and losers. They have a good idea when to sell though.

Stock markets are governed by rules and regulations. If some subpopulation of investors had an innate ability that gave them an extraordinary advantage over "normal" investors, then the markets would quickly adopt rules that would neutralize those unfair advantages. After all, nobody would invest in a market if they knew they had a massive disadvantage.

Precognition is advantageous in investing because it allows someone to trade with information not yet available to others. We have a form of this in the real world. When the New York Stock Exchange (for example) sends out information about changes in stock prices, companies located close to the stock exchange receive that transmission slightly sooner than companies located farther away. These timing differences are extremely minute, but significant enough in the high-frequency trading world that an advantage of a millisecond can mean that you can act on a market change before your competitor is even aware that it happened.

The difference between this and your precog world is that instead of requiring expensive high-speed computer and communication systems to game the markets, you have people that can do it with their minds. In the end, the same solutions that have been proposed to combat high-speed trading gimmicks would also work against your precogs. Markets wouldn't release updated pricing information in real time, they'd delay updates and release them in batches at pre-determined intervals. Instead of processing orders in the strict order they arrive, they'd consider all orders received within a certain window as "simultaneous" and process them in a random order. These speed bumps have to be long enough to equalize the differences between a "slow" and a "fast" actor. In the high-speed trading world, that might only be measured in fractions of seconds. In your precog world, that's more like 24 hours.

Also, consider that modern technology significantly reduces the usefulness of such short-term precognition. The vast majority of precogs can see less than a minute into the future. If they foresee a market movement worth acting on, they still have to execute the trade (input ticker symbols and prices, click submit buttons, etc). The speed of this process is limited by how fast a human can type, process visual information, and use a computer interface. High-speed trading algorithms can do tens of thousands of trades in time periods best measured in nanoseconds. Your "meatbag latency" is going to naturally negate the advantages for all but the most skilled precogs.

Stock Markets solve an information problem

Stock markets are a way we have of solving a pricing problem. Pricing problems are literally "what price should something be". It turns out that properly solving the pricing problem results in a huge amount of efficiency.

Precogs solve a category of information problems faster and better than anyone else.

Banning them might occur, but this is akin to Japan banning firearms. It is great because it keeps the people in charge in charge longer, at least for the people in charge, but it is a bad plan because other people who don't ban them end up showing up with big guns and forcing you to obey.

Given a problem in NP and O(2^N) computers, a precog with O(N) foresight can pick which computer will solve the problem before the computers run, resulting in an O(2^N) parallelism increase.

What more, if 2^N isn't large enough, the precog can respond with "none of the above", and they use that information to pick 2^N different choices, which then the precog can predict the result of. As the precog cannot predict the result of their own actions, until they act (and say "none of the above") the precog cannot predict the results of that.

If you have a 2nd precog that can predict the results of the first precog's actions to a high probability, you can get another exponential blowup in the power of precognition.

Suppose we have A and B. B can predict the result of A's choices 99.9% of the time.

A has a precog window of 1 hour.

You take a NP-style problem, each of whose sub-problems can be solved in 30 minutes, and you ask A to press a red or green button (green if in an hour there is a solution, red if not) as fast as they can. If they saw a solution and pressed red, they can go back and press green.

Every red button press they change which set of sub-problems they start solving. On a green, they back up, and they start a new process (where A quickly points out which sub-solution is the right one).

This generates a 1000x+ speedup. Not great.

Now we get B who can predict A's actions reliably. So B knows the result of A's chain of button presses over the next half-hour.

A then picks a number from a random number generator that selects the set of problems they are going to solve over the next half hour. B pressed a red or green button to say if, having picked a number, will A press the green button in the next half hour.

This gives you another 1000x+ speedup.

If you find C who can predict B and A, you can get another 1000x speedup.

The speedup factor grows exponentially with the number of precogs you can predict, and linearly with each of the precogs prediction window lengths.

And if you cannot work out a way to economically exploit 10^20 fold increases in solving pretty general classes of mathematical problem, you aren't imagining hard enough.

Precognition, of course, is much stronger than even this.

Companies using precogs would assembly lines that are retroactively safe. They have quality control that is retroactively perfect. A system design engineer could do the math, but completely different kinds of control systems can be used for scheduling, flight control, inventory management, etc.

Stick a precog in front of a fire reporting system, and we have fire trucks arriving before fires start, police arriving before crimes occur, bridges evacuated before they collapse, paperwork processed before person knows they need it, parts shipped before car breaks down.

Just in time inventory management becomes before-time inventory management.

The stock market, indirectly, goes crazy with massive returns. Money flows into whomever predicts the future best; people have to respond to future crashes now if they want to stay in the game, but faking a future crash doesn't work against a longer-duration/more reliable precog. Whomever (a) is playing that game, and (b) is confident they are the best, ends up attracting all of the money.

Slow stock markets might take off, with transaction speeds that are glacial. As the value of fast-pricing becomes lower and the profits to the few who can do it get higher, they may tax the profits of fast-pricing (before-day-trading) highly. Enough that messages get through (so it is worth it to precog and pass on the information), but "all the money" doesn't accumulate in the hands of the best precog.

• "In general, most precogs are short range (80% have a max range under a minute...)" Less than a minute isn't going to help with fires or most crimes, or much of anything where the actual time it takes to respond is significantly longer. Jul 23, 2019 at 20:23
• @KeithMorrison 1% of precogs, so a dozen or more people, have more than a day of precognition. And, as noted, you can do some work to chain under a minute precogs by having them predict (with some reliability) what another precog will do, who in turn is doing precognition. Algorithms for converting unreliable oracles into (more) reliable ones are well known in computer science, depending on exactly how they are unreliable.
– Yakk
Jul 23, 2019 at 20:27
• and why do you assume they'd want to or have any interest in doing the things you indicated? I know people who display an immense inherent talent for a given job, yet have utterly no interest in it. Jul 23, 2019 at 20:29
• @KeithMorrison You seem to be projecting. I didn't assume any one precog would be interested in anything. There is a dozen day+ precogs, 100+ hour+, and 1000+ minute+ precogs by the OP. Finding a use for their unique talent does not imply that all of them are doing it? Are you asking why we'd be able to get any of them to do it? There are piles of economic systems, and all of them are reasonably good (not perfect, just reasonably good) at persuading people to do useful things. That seems off-topic.
– Yakk
Jul 23, 2019 at 20:31

Background

The stock market is reactive to who is buying and who is selling. In today's exchanges, prices are modified by the second because of this concept. The price you see on a ticker board is essentially the price of the stock during it's last trade, which effectively becomes the basis for pricing future stock sales.

If there's a huge amount of sellers for stock X and not many buyers, the price tends to go down due to lack of interest. This isn't set by a computer, but rather by the seller, trying to sell off their shares to the "highest bidder", and the way that happens is by lowering your price until someone buys it.

If there's a stock Y that people really want because they support it or it's making good returns (or whatever), then the price people offer for the stocks make the price higher.

Granted, there's some regulation of the prices, but it's mostly about "supply and demand".

"Share prices are set by supply and demand in the market as buyers and sellers place orders. Order flow and bid-ask spreads are often maintained by specialists or market makers to ensure an orderly and fair market."

https://www.investopedia.com/articles/investing/082614/how-stock-market-works.asp

So what happens when you seriously change the conditions of a stock?

Any change to the supply or demand will cause the price to change. The problem with seeing the future is if the changes the pregoc makes actually increase, decrease, or cause the expected market value of a stock.

For example: if a precog sees a 20% increase in a stock in one day and they buy 10k shares, are they the one that caused the price increase? Does the stock value instead go up by 30%? Does the price instead drop by 10%, because everyone is now trying to sell due to the massive purchase? And can your precog see the change in time to change their purchase from 10k shares up to 15k or down to 2k, depending on the desired outcome? How far in the future can the see? Is it just a few seconds, minutes, days, weeks? And does that change per person, and is variable based on skill or static based on natural talent?

A massive change to a stocks value or trade volume can cause other people to take notice. If enough people, or a large individual investor, follow along with the new trend, it can also affect the valuation of the stock, positively or negatively. This happens without precognition and is called a "run". Either people are running to the stock or away from it. If the run is bad enough, the people monitoring the exchange can cease trading that stock to prevent serious market problems. These problems can kill a company if they are serious enough. They can also seriously affect the future of the stock even after the run ends.

So what happens when you have someone who can predict the future?

There's a couple things that can happen.

A careful/moral investor will make sure their decisions don't seriously affect the expected price by their interest in that stock. They will purchase and sell modest amounts of stock, maybe through intermediaries, and try to prevent runs that cause havoc. They notice an expected rise or fall in prices and invest accordingly.

An immoral investor will try to game the system. They will purposefully cause the stock to be noticed and try to get the price to rise unnaturally, then sell before trading stops or the price takes a nose dive. They will notice a natural rise or fall in prices, then make massive purchases or sales in an effort to make it more drastic and to their advantage.

A newbie may unintentionally cause stock prices to unnaturally rise and fall, but will simply be inexperienced on how to prevent this. They may lose as much money as they earn. They may also lose considerably more than they earn, simply by not paying attention.

Then again, that I've explained happens in today's market. The only difference is that we currently work on historical trends instead of future/expected trends.

There's a problem... and complications...

Unless your Universe is pre-destined to everything, then current decisions affect the future. If you've ever seen the movie "Next" with Nicholas Cage, you'll know that a precog might be able to "try" different things to cause an expected outcome.

In the movie, Cage is able to project himself into the future a variable amount of time, seeing what happens, then changing how he progresses. This is useful to him as he can now predict when and where he is shot, in one scene, so he can effectively dodge the bullets before they are fired. It may not be the best movie around, but it does make you think about how effective being able to see into the future may be, or not, depending on your intelligence.

Are your precogs able to "see" the changes they make in the timeline in "realtime" or are they effectively as blind as anyone else to their purchases and sales. If they see a price go up or down and they make a significant purchase or sale, does that price they see match the price at the time actually forecast? Asked differently: if they look again, does their "vision" change? Can they even look again or is that same instant in time blocked from them somehow? Is it only a function of energy whether they can look forward and how long is their recharge time? This kind of meshes with how far into the future they can see, because if their cooldown is longer than their pre-vision, they might not be able to change their decision anyway, unless there's partners or a team making sure the decisions are correct.

Conclusion

So, does your stock purchase change the future? Can you "ride the wave" of someone else's run and get out before you get stuck? Or does one of your characters realize that precog doesn't change anything and therefore doesn't really matter (pre-destiny)? Maybe there's another set of precogs that work to prevent runs on stocks and effectively "normalizes" the market? Maybe the market is regulated so that only a certain amount/type/size of unnatural market fluctuation are made before a specific individual is required to get more education or is outright banned.

This is your story, so you have to decide what what you want to happen. If you want to make the stock so dangerous that only a fool or genius would participate, you can do that. If you want it to be so regulated that only "masters" allowed to use real money, with everyone else being on a virtual market until they become "certified", you can do that, too. Maybe the stock market was so volatile that it crashed years ago, and no there's only stories/myths about it and how people went mad trying to use it.

Hopefully I provided more insight than questions. "Have fun storming the castle!"

They would totally work in arbitrage with total approval of the government --- reducing volatility, they lower the average cost of borrowing across the market and thus provide a useful service.

I was always abhorred by the idea of arbitrage (literally making money from nothing; boils down to having an internet connection and hardware that's a few milliseconds or fraction thereof faster than the rest), but beyond that it has a stabilizing effect.