Adaptive Shortfall


The aim of Adaptive Shortfall Strategy is to minimize the market impact of orders execution and therefore reduce cost of executing large orders with averse of time risk.

As in casual Implementation Shortfall, adaptive one works in some time span or optimized Trade Horizon. It trades depending on asset’s historical volume changes. Strategy divides trading horizon into portions (intervals) in which tries to obtain minimal market volume participation essential to achieve demanded security bought/sold volume in given time. Basing on historical market volume changes, it predicts proper volume participation for current interval and tries to trade it. Moreover, execution of every interval might be fulfilled with any successive methodology, like TWAP or POV. The Trade Horizon is indicator of risk aversion – it may be passed to strategy as it is or calculated automatically basing on asset’s parameters and risk aversion scale.

The measure of strategy’s effectiveness is comparation to some Price Benchmark (which may be not defined and then substituted with another indicator, like the midpoint price at the moment). The less is the gap – the more efficient is an algorithm’s execution.

Algorithm being described is adaptational in regard to asset’s price. According to Strategy Bias, it can be aggressive in-the-money (AIM) or ‘’passive in-the-money’’ (PIM). Bias set to AIM causes strategy to take adventage from favourable prices and buy more security when its price is below the benchmark one (or sell when it is above). Bias set to PIM is actually the opposite and makes strategy to buy faster when asset’s price rises above benchmark (or sell faster when it drops). AIM is generally more efficient during price reversals whereas PIM is a good strategy for consequent trends. Strategy Bias influence can be fluently adjusted. If it is not given – strategy occurs like a standard Implementation Shortfall, ignoring price dynamics.

Furthermore, like all casual Shortfall strategies, Adaptive Shortfall can be tilted. To reduce time risk strategy’s trading in beginning intervals is more aggresive, then market participation decreases slightly with time. Also, stop loss mechanism can be implemented to automatically prevent strategy from executing inappropriate trades during critical market conditions.

Market Data

  • Last trade
  • Order book
  • Historical market data


Trade Horizon Required maximum time of strategy’s active trading (considered as Risk Aversion) Yes
Volume Required strategy’s overall traded volume Yes
Price Benchmark Price for determining strategy’s effectiveness No
Limits Absolute limits of orders’ prices No
Interval Length Length of the single trading interval Yes
Strategy Bias Defines if strategy is Aggressive In-the-Money (AIM) or Passive In-the-Money (PIM) No
Tilt Indicates whether strategy should be more active in the beginning or not No


Open position

Side Buy or Sell
Amount Part of Volume adequate to current interval market participation
Price Last market price
Type Price limit / Market order

Adaptive Shortfall strategy’s opened positions strongly depend on subalgorithm used in implementation. It keeps calculated market participation in each interval but this participation can be approached by other algorithm, designed for it.

Close position

In general, strategy does not open more position than it should execute in total. So, if the stop loss mechanism does not have to be used, there is no need to close any in other way than executing them.


Strategy terminates within given Trade Horizon, after execution all of demanded Volume.

Time frame

Strategy bases on some Trade Horizon and its main aim is to reduce market impact during quick, usually one-day executions of very large orders. So it is generally designed for daily trading. Some initial market volume data collection may be essential before trading if such information is not provided to the startegy.

Further information & source

  • Barry Johnson, Algorithmic Trading & DMA: An introduction to direct access trading strategies, 4Myeloma Press, 2010.