Wall Street Built a $370B Business Cloning Quant Trades | Hacker News Reader

Wall Street Built a $370B Business Cloning Quant Trades

bloomberg.com

28 pointsmfiguiere10 comments

Open articleView on HN

https://ghostarchive.org/archive/9bFYU

you da real MVP

I like this

I don’t believe it’s as simple as the articles puts it.

How is the alpha in this strats not exhausted? Either they don’t make money or it’s not fully transparent.

Yeah the example they give "say, go long S&P 500 futures if they’ve risen above a 20-day moving average, in one simplification." pretty much goes against everything I have read about quant strategies. They all say that technical analysis isn't useful.

If this is the strategies they are selling it seems like a cash grab, and not an alpha producing strategy.

> pretty much goes against everything I have read about quant strategies. They all say that technical analysis isn't useful.

could you expand on this? I always understood it as, technical analysis has become a self fulfilling prophecy and you have different players trying to manipulate that fact

you've got pressures (sizable... I think? not really sure how to measure) from ETF inflows long + short from the "day trading/swing trading" crowd, also the "0-day-to-expiration SPY/QQQ options" gambling crowd.

Then you have the "hedge funds" with the "high frequency trading algorithms" sitting there in a predatory fashion trying to manipulate the emotions.

"Just because the 20/50 SMA crossed doesn't mean it's going to rocket because I'm going to pray on the opposite"

That's how I see at least, curious to hear if I'm off my rocker. I agree with you, no mixture of boiling down the market into simple signals off of things like volume/patterns/averages is long-term exploitable, and that's the whole problem. What works sometimes doesn't work others because of all of the different factors/players at play.

With that said, is there any conclusive research on whether algorithm/quant based trading is actually profitable (generates alpha)?

Not sure if any studies were done, but you need go no further than the folks who invented it, Jim Simons and Renaissance Technologies, whose Medallion Fund returned 66% over 30 years. The fact that they closed the fund to everyone but employees soon after they realized how good their algorithms were says a lot.

https://en.wikipedia.org/wiki/Renaissance_Technologies

Bloomberg is just trash at this point when it comes to articles. Of course no one would consider that a quant strategy besides a Bloomberg reporter. That is hardly surprising to me though.

This is a product that is sold. That should tell you enough. That is, this stuff doesn't contain any alpha

Yes, that would make the most sense, but why funds buy it? Are they really that oblivious?

Reply on news.ycombinator.com