We have noticed an uptick in market volatility, leading traders to rely more heavily on VWAP algos for execution. While these algos generally perform well on average, we've identified specific challenges worth addressing:
Variability in Trade Outcomes:
Individual trade outcomes with VWAP algos can significantly deviate from the benchmark. We have seen some datasets where over half of the trades deviate by more than four basis points, predominantly on the negative side.
Implementation Shortfall Issue:
Many TCA providers tend to downplay the issue by focusing solely on comparing VWAP trades to the VWAP benchmark, thus obscuring the full scope of slippage. While news is used as a way to “explain away" outliers, this is a mistake as poorly designed algorithms will incur substantially more slippage on news days.
Impact on Event Days:
VWAP algos struggle on days with anomalous market conditions, often resulting in deviation from the volume curve and VWAP.
We have heard traders tell us that they don’t care if they miss the VWAP because they are using the algo to hedge against an unpredictable trend. One of the faults in this logic is that VWAP algos are backloaded to execute more at the close. This will result in significant underperformance on days when there is an adverse trend, which is also when they are most widely used.
“Cost-Saving” Mechanisms Are Costly:
Some VWAP algos employ “cost saving” mechanisms that can inadvertently exacerbate cost in trending markets. Pattern-masking techniques can result in small fill sizes that create opportunity cost in a fast market, but the biggest issues are measures that drive routes away from liquid primary exchanges and cause the algo to fall behind, such are avoiding exchanges because of rebates and attempts to minimize reversion.
Assuming that non-primary and dark pool liquidity is universally "less costly" than shares on a primary exchange is also flawed logic. In cases where price is moving unfavorably, falling behind in volume due to insufficient liquidity can lead the algo to "catch up" at worse prices. This often results in most VWAP algos completing the order at or near the day's worst price by day-end.
The overriding goal for the trading process is to preserve portfolio value. Given that generally all providers are meeting the VWAP benchmark in aggregate, why use an algo that can move the price more? Some tips to help you keep your eye on the performance prize and reduce that fat tail:
Prioritize primary exchanges in trending markets.
Consider POV when price is trending against you.
Evaluate trades against both VWAP and arrival—and keep an eye on fat tails.
Understand what the most liquid venues are and aren’t for each order that arrives on your blotter. Match this order and the prevailing and expected market conditions to the appropriate algo.
