After removing fake trades, does volume still predict price moves, and how much?
Research question: Do the volume-return interaction results in [2] replicate in samples that use wash-trade-filtered volume, and how much of the predictive power remains after filtering?
The records do not show whether the volume-return interaction results in 2 replicate in wash-trade-filtered samples, because 2 reports only that a Columbia University analysis found roughly 25% of Polymarket's historical volume showed signs of wash trading and does not address return predictability. The closest related evidence is that lagged cryptocurrency returns positively predict wash trade volume 3, but that paper does not test replication of 2 or report predictive power after filtering 3. Filtering does affect performance in related work, as the standard LAMV portfolio's Sharpe Ratio falls from 1.41 without wash-trading treatment to 0.96 with it 1. Treatment removes essentially all small-volume wash trades but only some large-volume ones 7, so the share of predictive power that remains cannot be determined from these facts. The newest fact is dated April 2, 2026 2, about six months before October 2026.
Sources
"We notice that the standard LAMV portfolio with treatment on wash trading (Portfolio 8) has a lower SR (0.96) than that from the standard TMV Portfolio 7 (SR=1.35) and the standard LAMV Portfolio 9 without treatment on wash trading (SR=1.41)."Liquidity Jump, Liquidity Diffusion, and Treatment on Wash Trading of Crypto Assets
"A Columbia University analysis of Polymarket found roughly 25% of historical volume showed signs of wash trading"Crypto’s wash trading problem is ‘far more common’ than investors think, DOJ sting shows
"In Table 10, lagged cryptocurrency returns positively predict wash trade volume, while lagged volatility shows a strong negative prediction."NBER WORKING PAPER SERIES CRYPTO WASH TRADING Lin William Cong Xi Li Ke Tang
"Moreover, wash trading is positively predicted by returns and negatively by price volatility."NBER WORKING PAPER SERIES CRYPTO WASH TRADING Lin William Cong Xi Li Ke Tang
"Price increases could draw retail investors’ attention and encourage speculation. Therefore, crypto exchanges are incentivized to pump up volumes to vie for better ranking and more clients."NBER WORKING PAPER SERIES CRYPTO WASH TRADING Lin William Cong Xi Li Ke Tang
"We demonstrate that treatment of washing trading significantly reduces the liquidity diffusion, but only reduces the liquidity jump to a level that is inadequate to restore the asset’s responsiveness to the autoregressive models."Liquidity Jump, Liquidity Diffusion, and Treatment on Wash Trading of Crypto Assets (2024)
"Our results indicate that treatment on wash trading essentially removes all small-volume wash trades, but only some large-volume wash trades."Liquidity Jump, Liquidity Diffusion, and Treatment on Wash Trading of Crypto Assets (2024)
"We argue that treatment on wash trading is unnecessary in modeling established crypto assets that trade in unregulated but mainstream exchanges."Liquidity Jump, Liquidity Diffusion, and Treatment on Wash Trading of Crypto Assets (2024)