quadratic voting
What is your best guess of the strategies that Renaissance tech use?

does Medallion provided audited financials? No (because it is closed to outside investors so not required). Performance? awesome.
do other RenTech funds provide audited financials? Yes.
Performance? mediocre
HmmmJust hypothetically though, to get back to it. Earning 50% return consistently is extraordinary.
How would you even be able to do it without having a few real bad losses? It's the consistency that is also incredibly.

Good data collection for decades. Most accurate datasets. Basically a great IT team.
Great management that understands the scientific process very well and commands the respect of a workforce of highly technical, arrogant and difficult to manage STEM PhDs.
Incentives are aligned properly with the employees owning equity in the private fund.Good governance, incentive structure and technology.

What I've heard is that they have some strategies with modest returns but _very_ low volatility, and then lever up like crazy.
does Medallion provided audited financials? No (because it is closed to outside investors so not required). Performance? awesome.
do other RenTech funds provide audited financials? Yes.
Performance? mediocre
HmmmJust hypothetically though, to get back to it. Earning 50% return consistently is extraordinary.
How would you even be able to do it without having a few real bad losses? It's the consistency that is also incredibly. 
They find a million small edges in the market.
Then they have one giant model that takes all of those edges, computes the correlations, figures out which to apply, when, what leverage etc, and then they trade all of them at once.
The trick is they basically still have alpha research, but they can find strategies that only making like 50k a year, but they find a million of them. The model they have makes sure the strategies don't clash. This uses exotic and advanced statistics.
An anecdote:
I used to know a market maker who said it was common to see orders coming in for months before a stocked popped huge. He said they tried to hide their buying using all sorts of algorithms, but that using kalman filters on market orders, he could uncover consistent buying patterns for specific stocks. Using that, he found out that the buyer was often renaissance (not sure how he found this out). And so would look for those buys and then piggy back on.

Inside trading
Traditionally, wall street has been all about insider trading. The "analyst" who went to Harvard is buddy with the CEO at some corp. They go out drinking and discuss "innovation", buddy finds out if the company is a buy or not. And voila he collects his 2 million a year pay check.

Rentech was a loser for a few years at the beginning too, and took some work to get “right.” If they have a secret sauce, it’s a broad series of processes that really effectively matches a trading strategy to probability theory. No one has figured out what those are though.
so why do rentech do so much better than say aqr? given lots of smart people at both.
AQR has been massively expanding in recent years (both in personnel and asset under management). doesnt make sense to compare rentech and AQR currently, until we have another 5 years at least.
these are both examples of success, so far.
If you want examples of bad funds, plenty you can name.
Probability THEORY doesn't really help if you don't know the real probabilities and what impact them. What's the probability that Tesla will go up $100 by tomorrow? What do you look at?

they can not be so successful if you can guess what they are doing. No.1 rule in hedge fund industry is: don't talk.
Markets are efficient in the long run, but they are locally very inelastic and present patterns all the time. It's not that hard to find 4/5 uncorrelated short term strategies that give you a success ratio of 55/60%. With that you can generate very high Sharpe Ratios.
Just to elaborate. Imagine you have 5 uncorrelated bets that win 1% w.p. 0.55 and lose 1% w.p. 0.45. The expected return is 0.1% with standard deviation of 1%. The 1/N portfolio has standard deviation of 0.44%. The annual return is 252*0.1% = 25.2%, the standard devation 7%, the Sharpe ratio (25.20)/7 = 3.6!
If the success ratio is 60% you get annual expected returns of 37.8% and a Sharpe ratio of 5.55!55% hit ratio is hard. especially oos.
Not as crazy as you think. All the strategies that provide liquidity (like short term reversal) have much better odds

Win every day matched with a bunch of guys with a deep understanding of esoteric probability theory. At least at the Medallion. Find
It’s no coincidence the dude hires astro physicists.
That said, i bet it’s mostly luck.Hard to believe in luck when they have made 66% average return consistently beating the market
https://www.institutionalinvestor.com/article/b1l98yt4p0bvr4/TheFamedMedallionFundIsCrushingItOtherRenTechFundsNotSoMuch
Simons himself has said they rely in very short lived signals.Awww how cute. Unfortunately the real world doesn't work that way.
Early on (way back), Simons took basic technical indicators and programmed a computer to do the analysis and work. That's where the name "Renaissance technology" comes from. It paid off pretty big.
Since then, it's been insider trading.