Why Reversing Losing Trader Flow Does Not Guarantee Profit

Summary

The document considers whether a trader can profit simply by taking the opposite side of a stream generated by traders who lose money. Its answer emphasizes that observed losses are typically measured after transaction costs. A candidate group must lose more than the costs the contrarian trader will also pay, and aggregate evidence that many traders lose does not establish that a selected subset has predictive value beyond noise.

Even a statistically meaningful edge would not settle the trading question. Returns may depend on how those traders size positions; for example, countering leveraged averaging down can resemble a trend-following exposure, with losses when prices reverse. The response also distinguishes broker economics from directional prediction: banks may earn spreads on customer flow without assuming customers are wrong. The discussion gives qualitative cautions rather than a tested strategy, sample, or performance estimate. It does not specify how to identify losing traders, estimate their future order flow, or manage execution and risk, so profitability cannot be inferred from the premise alone.

Key ideas

• Trader losses after fees do not automatically create a profitable signal for someone trading against them.
• The selected order flow must show losses beyond the costs the contrarian strategy will incur.
• A statistically significant edge can still be sensitive to client position sizing and reversals.
• Countering averaging down may create trend-following exposure with meaningful reversal risk.
• A broker can earn from spreads without relying on a belief that its customers predict prices poorly.

Tags

# Trading against a loser flow


# Trading against a loser flow












Let's say we are getting a trade stream(instantaneous) of a group of traders that lose money. If we trade on their opposite side in a broker, is it guaranteed that we make money? Intuitionally it seems possible but it may not be that simple... Any comment is appreciated..

## Answer by Lliane (score 2)

https://quant.stackexchange.com/a/43701

There are a couple constraints :

- It is true that the majority of traders are losers, but that's after fees. You need to find losers who are statistically losing more than just the fees (because you're going to pay the same), fees can be a sizable share of the losses in the long term. There is no reason to think that the subset of order flow you isolate is any different from random noise, just like Attack68 mentionned above.

- Then even if you manage to find a statistically significant edge, you become dependent on the sizing of those clients, if you assume for instance that the average retail loser is just a guy averaging down on leverage until he busts, the opposite strategy is a CTA momentum/trend following, you'll need to be confident with the sizing that comes with it (reversals will hurt).

It's been mentionned that banks do that against their FX customers, but the way they get paid is on the spread on their order flow, not because they think their customer is wrong (it might also be because they manipulate the fixing, but that's not legal and anyway that's not something you can do).

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.