Framing the Exit Order for Multiple Open Trades

Summary

The document poses a trade management problem: how to choose the order for closing a portfolio of open long and short positions when some positions hedge one another and others leave net exposure. The positions have been open for different periods and are subject to stochastic price changes. The author asks whether the decision can be framed as an optimization problem involving inventory risk, profit, or another objective.

One tentative rule is to close the largest winner when cumulative profit is positive, or the smallest loser when it is negative. The author doubts that this rule is useful and asks whether a principled optimal strategy exists. The supplied text contains no answer, model, execution constraints, or evidence comparing exit policies. It therefore identifies a general problem rather than offering a tested method; any solution would need to specify objectives, risk preferences, trading costs, and how closing one position changes the remaining portfolio’s exposure.

Key ideas

• The order of exits can matter when open positions are only partially hedged.
• The problem can be posed in terms of inventory risk or profit maximization.
• The proposed winner and loser rule is presented as tentative, not as a tested strategy.
• The document does not provide an optimization method or evidence comparing exit policies.
• A practical objective would need to account for portfolio exposure, trading costs, and risk preferences.

Tags

# How to "best" exit multiple trades?


# How to "best" exit multiple trades?












Let say I have N opened trades (N = s + b) that are partialy hedged, and paritaly not. In general s != b. Some of them are market sell orders (s), and the rest of them are market buy orders (b). They were all opened some time ago and currently they experience stochastic value change according to price dynamics.

I would like to close them one by one using... some optimal strategy. Can this problem be formulated in means of inventory holding risk, profit maximization or any other useful metric?

My first thought was to exit largest winner when cumulative profit is positive, and exit least looser when cumulative profit is negative - but is this realy useful??? (now i think it's not)

Please give me some advice. Can this be done optimally?

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.