A financial option is a rather complex instrument. Unlike delta-one products, an option value depends not only on the underlying, but also on volatility, time to maturity, strike, interest rate, and dividends. Options have been used as hedging instruments, but they’re becoming a speculative vehicle these days thanks to a growing number of retail traders and the increasing popularity of financial media.
Even though options are derivatives, and thus trading them would require a different thinking, retail traders often apply delta-one trading techniques to options trading. One such popular technique is stop loss. The trading rule usually goes like this: sell options to collect premium, and if the mark-to-market loss exceeds a certain multiplier of the premium received, then exit. On this topic, Reference  examined the role of stop losses in options trading. Specifically,
This paper is devoted to the research and development of profitable option sell-side trading strategies, and proposes an operating mechanism for stop-loss. In addition, statistical methods and random forest algorithms are used to estimate the win-rate of the strategy. The win-rate represents the proportion of all transactions that the premium has not doubled before settlement, and we can also express it with precision.
After performing numerical experiments, the authors concluded,
The experimental results can confirm that the trading strategy proposed by this paper can effectively achieve risk control through the development of a stop-loss mechanism with a fixed premium double multiple. And apply statistical methods and random forest algorithm to estimate the win-rate of the strategy, and screen out the trading range with higher profit and stable. The precision predicted by the model classification can prove that the strategy is practical and profitable.
The paper led us to ask ourselves the following questions:
- Why do we use a stop loss instead of employing a defined payoff position, e.g. a vertical? Recall that an option position is a bet on either the underlying’s terminal distribution or its dynamics. Using verticals would allow us to have a well-defined bet on the terminal distribution of the underlying. If we sell an option and then apply a stop loss, what the nature of the bet would be?
- The use of stop losses would prevent us from taking advantage of the mean reversion property of stock indices. Defined payoff positions will allow us to do this.
- Can we get realistic fills using stop losses in options trading? It’s well known that, even in a liquid, primary market, when volatility increases, stop losses will become less effective, i.e. traders will find it more difficult getting fills. Options are less liquid, derivative market with bigger Bid-Ask spreads. Can stop losses be applied effectively?
These are just some questions. Let us know what you think.
Last and not least, we also observed from the paper:
- Figure 1 is incorrect in the context of time decay in the real world.
- Predicting the win rate is meaningless if the strategy’s expectation value is negative.
 C-F Chao, Y-C Wang, M-E Wu, A Quantitative Model for Option Sell-Side Trading with Stop-Loss Mechanism by Using Random Forest, 2021, https://www.researchsquare.com/article/rs-769898/v1