AI and Bitcoin are driving the next big hedge fund wave | Hacker News Reader

AI and Bitcoin are driving the next big hedge fund wave

wired.com

80 pointsoska35 comments

My terse and cynical reaction is, "Good luck with that." Of course, some people will become very rich, but just flipping coins gives you a bell curve.

It reminded me of "thinking fast and slow" book where the author refers to the study conducted by Terry Odean and Brad Barber:

"In a paper titled “Trading Is Hazardous to Your Wealth,” they showed that, on average, the most active traders had the poorest results, while the investors who traded the least earned the highest returns. In another paper, titled “Boys Will Be Boys,” they showed that men acted on their useless ideas significantly more often than women, and that as a result women achieved better investment results than men."

and concludes:

"It is important to remember that this is a statement about averages: some individuals did much better, others did much worse. However, it is clear that for the large majority of individual investors, taking a shower and doing nothing would have been a better policy than implementing the ideas that came to their minds."

Kahneman, Daniel (2011-11-03). Thinking, Fast and Slow (p. 214). Penguin Books Ltd. Kindle Edition.

As a small personal trader you're probably paying a non trivial amount in bp space for each move in and out of a position (whilst competing with companies which don't).

The average actively trading investor will inevitably lose money vs the underlying benchmark index(es) that represent their portfolio.

The secret with building good models is and will always be about getting good data. There is a ton of good data for trading. Unfortunately, the SEC frowns on it being used for profit.

> Unfortunately, the SEC frowns on it being used for profit.

That's not true. You can use as much good data as you want for trading. As long as it doesn't come from the officers of a company (directly or laundered) before that data is announced publicly.

Yep, that is where I was going.

There are no cheap options on Bitcoin. There is no short selling of expensive options even if they exist. Every time I see hedging with Bitcoin in the same sentence, I smell fraud.

Not mentioned in the OP but there was also news out today that hedge funds are buying up Mt.Gox creditor claims.

https://www.cryptocoinsnews.com/report-hedge-funds-buying-cu...

That's an interesting trade. If you entirely ignore the Bitcoin aspect of it, you're just investing claims against a bankrupt Japanese company which is being competently administered in the ordinary course of a 民事再生。 Currently 100 yen worth of claims looks like it will receive 25 yen worth of satisfaction at a non-deterministic time between one and two years from now. But, but, but: something like 20 of those 25 yen are presently tied up in a non-liquid-but-very-volatile[+] security, so buying bankruptcy claims gets you effective 2:1 leverage on that security. (Handwavy assumptions: you're sure of getting the 5 yen back that is actually presently held in yen and you can borrow yen for approximately 0% in quantity, so essentially you're paying 10 yen for 20 yen worth of exposure to the security.)

[+] Yes, you can sell 1 BTC without meaningful price impact, but you can't sell 200k BTC without meaningful price impact, so it is not liquid by the standards of hedge funds.

That's just some bottom feeder buying up claims at a discount. There's a small industry which does that.[1] Sometimes the goal is to buy the business at a big discount and do something with the assets.

[1] https://www.law360.com/articles/498711/understanding-the-rea...

That was my first thought but, crucially, the bankruptcy administrator is not holding a pile of yen. If he were merely holding a pile of yen, this is a pretty boring deal, and you'd be right -- it's a straightforward swap of liquid cash for illiquid cash with an embedded premium for legal risk.

But the bankrupt entity holds 200k odd Bitcoins, and that changes things a lot. I wouldn't make this trade even at an 85% discount to face value, because I think there's a significant likelihood that BTC goes to zero, and so would not buy it and especially would not buy it with embedded leverage.

I assume you mean that you think there's a significant likelihood that BTC goes to zero before the BTC are returned to the creditors, right?

Subtly different: I think there exists a likelihood that BTC goes to zero before the BTC are sold. (BTC will not be returned to creditors directly, I predict with a fairly high degree of confidence. They'll be sold as a block; Kraken might be separately engaged to transfer money to foreign creditors who might take their transfers in BTC.)

When the article has "ai" AND "bitcoin" in its title, AND it's from wired, you know you're screwed.....

Actually, there's something to this I've been thinking about for a while...

Indeed this is happening. Pricing out the 250k+/yr quants as the field is levelled.

I'm surprised the hedge funds didn't figure out earlier that having everything proprietary would drive their costs so high. When there are so many players looking for alpha the only variable to control is costs.

Can't the same quants go make more money helping to train the bots?

Training bots is now for suckers.

It's all about training bots to train bots. Choo choo

No that's not the cause, globally fees are coming down across the industry and that is putting pressure on salaries.

This article contains so much misinformation and silly inaccurate hype that it's almost painful to read for those of us in this space. I mean even for Wired.

A lot of crude misinformation. But this is happening, it's not just hype. It does help getting people to flip over, though.

This could be interesting, but without pointing out what's wrong, or knowing your experience it reads like a vague military dismissal of an intelligence leak.

What's wrong with the article? This is not a dispute but a request for more information.

(With hesitance, that I'm just piling on the negatives: I actually tend to ignore people who say things in an authoritative way but make no effort to substantiate their claims, because more often than not they're just wrong. I would assume likewise here).

Hopefully not together in the same investment!

AI, probably. Bitcoin, probably not.

I know HN frowns on this kind of comment, but: suuuuuuure.

Numerai is generating an unprecedented amount of hype articles.

Exactly, I played Numerai many months, now it is full of cheating robots.

I'm curious, how do the robots cheat?

There are people submitting a ton of models that are all very similar, basically overfitting the public board.

How does that make sense? In an AI competition you expect robots to compete to the best of their ability.

The bots are used to scrape data from the leaderboard and make adjustments to improve their score by overfitting the public leaderboard data.

Essentially this:

http://blog.mrtz.org/2015/03/09/competition.html

You might say: but that won't change who actually wins on the private board. However when you use techniques like that you can get identify data in the test set and use it for training. It does help a some, but not a lot. The payout isn't very large for the over fitters (maybe 10-20% of payout).

Another negative is regular competitors also have no idea how they are actually doing compared to others. The leaderboard has impossible (for market data) logloss values of 0.2 - 0.4, sometimes even 0.00. They keep trying to fix it, but they haven't in about 2 months.

In this case, bots are automated cheating.

Even my cheating robot is bored of Numerai.

i'm guessing they're submarine'ing, maybe hired a PR firm

I wonder if they are actually making any money...

Reply on news.ycombinator.com