"The Strategy": a theory of market manipulation by big funds (Knuteson, 2018)

[d08]

Why don't you backtest your statement with SPY and see how profitable.

I did. Backtested SPY from 1993 to 2019: Bought at close price and sold at next morning's open price. Not that profitable especially not as good as buy and hold SPY from 1993-2019, even without subtracting the cost of bid/ask and commissions.

Look at it in relation to the drawdowns. Equalizing drawdowns, it offers much better returns I think. Not to mention that lower exposure is always lower risk.

 

[misterkel]

Both of his propositions would have had inferior returns as compared to “buy and hold”. D.E. Shaw, where he was a quantitative analyst, is a HF strategy firm AFAIK.

The problem with your claim: in the absence of such large players, the markets would not have moved upward. They are creating the 'buy and hold' market. (I think)

 

[bone]

The problem with your claim is that the large players are supposedly selling out their positions in the afternoon, which would put tremendous downward pressure on the market - which we haven’t seen in reality going into the close. And if many small retail traders are buying and holding - to date they have been handsomely rewarded.

The problem with your claim: in the absence of such large players, the markets would not have moved upward. They are creating the 'buy and hold' market. (I think)

 

[geneticien]

First, one should really consider the updated version of the paper https://arxiv.org/pdf/1912.01708.pdf since it is much clearer.

Then, the author does not say that the big players move the whole market. Thus, it does not make sense to invalidate his claims by providing some statistics about buyholding S&P 500.
Since only a subset of stocks is considered, the whole scheme is a pump&dump which, according to the author, runs for the last 30 years in an organized manner.

His explanation is of course simplified but it is sound.
As a side effect, Figure 2 and 3 explains why daytrading is unprofitable in the wast majority of cases (especially for retail traders) - because in expectation the price will dump within a single day.

Also, it is not specific to one country but according to the author a practice used worldwide.

Regarding his hints about big players. Perhaps, it is not only DE Shaw he worked for or RennTech. Also, it might be BlackRock which uses only a fraction of their AUM for active trading.

 

[d0rian]

Here's a slightly alternate version of the paper posted.
https://arxiv.org/pdf/1912.01708.pdf

I'll take a stab.

My initial takeaway, is that he is saying that markets can be favorably manipulated by large enough short term players that implicitly collude, and he explains some possible methods that they 'could' apply to achieve this. He argues that as long as the manipulation is accomplished in such a way, that the overall drift is beneficial to overall market players (who most likely desire positive drift), then very few watchdogs will be likely motivated to investigate the manipulation.

He starts by explaining one such mechanism that would benefit large players. Since the markets have a greater potential impact in the morning, a large player (or players colluding) could take a set of large positions in the morning and those positions would move favorably (on average) by some distance, call it x. If they take the exact opposite positions at the end of the day and close them out, the impact (movement) due to their closing out of those positions would be smaller, so that the net return would be positive because of the asymmetry of impacts at different times.
Over time the cumulative net impact of these actions benefit those firms that have the buying power to move markets using this type of scheme.

He goes on to argue that a more complicated favorable manipulation scheme could be applied at different times in the day, that might explain the overnight effects patterns across global markets.

I know people get very passionate about manipulation topics, so I won't go any further. But that's a casual interpretation of his paper(s).

You appear to be one of the few who actually read (or understood) the paper(s); this is more or less what I'm getting from them too having now re-read them a couple times.

 

[d0rian]

I've re-read both of his papers (1, 2) a couple times now since this discussion didn't seem to be going anywhere productive. It doesn't appear that Knuteson is the quack / bitter retail trader with an axe to grind many ITT seem to believe him to be -- he's not pitching Martingale 2.0 / "here's how I got rich, now you can too!" hokum; simply offering a theory for what might account for the well-documented day/night pattern of returns over the last 30 years (with nearly all gains coming outside regular trading hours). The 2nd paper, as a few others have pointed out, does a much better job describing the mechanics of the (allegedly-manipulative) trading practice; TLDR:

i) Amass a large basket of equities ($1B+)
ii) Use a smaller chunk of your fund to buy those equities in the morning when spreads are wider and volatility higher, and sell them back near close. The a.m. trading will move the market more than the p.m. closing out because of the higher liquidity.
iii) It's not a day-trading / open positions in the morning >> close at night thing -- those ITT saying as much clearly didn't read the paper. It's a buy-and-hold strategy, with an asterisk: a small chunk of the investment is used to engage in complex intraday round-tripping that has the effect of nudging the market upwards. They lose money on the smaller intraday chunk, but (per author's theory) more than make up for it with the market gains to the larger portfolio.
iv) Author claims that this theorized trading activity is the only one put forth to date that can account for the day/night return pattern. Yes, major news / earnings are announced after-hours, but studies isolating for those events suggest they only account for a minority (10-20%) of the day/night disparity.
v) Author claims that by the strict letter of the law, this trading practice is illegal market manipulation; but because its effect is to nudge the entire market upwards, no one complains too loudly because everyone, including those whose job it is to look into this stuff, is happy making money with the rising market and doesn't want to rock the boat.
​

 

[geneticien]

the question is whether one can detect the strategy, i.e., which stocks it is dealing with at the moment