Choosing EMA Inputs for Options MACD: Underlying, Option Price, or Implied Volatility
Summary
The discussion asks whether MACD and its fast and slow exponential moving averages should be calculated from an option’s market price or its underlying. One response argues that smoothing the underlying before using it to calculate an option premium would lower the estimated underlying volatility and therefore the computed Black–Scholes premium; for a display of market prices, it recommends applying the EMA to the option price itself.
A second response distinguishes the purpose of the signal: technical indicators such as MACD are generally more suitable for the underlying than for option returns. For a volatility-focused trade, it suggests that MACD on implied volatility could be relevant. These replies do not establish a universal choice or provide tests of signal performance. The appropriate input depends on whether the analysis targets the option’s traded price, the underlying’s directional movement, or implied volatility, and the document offers no evidence that any proposed MACD variant is profitable.
Key ideas
- Applying an EMA to the underlying and applying it to the option price answer different analytical questions.
- Smoothing the underlying can reduce estimated volatility and lower a Black–Scholes premium calculation.
- The replies generally favor underlying-price indicators for directional signals and raise implied volatility as a possible input for volatility trades.
- The discussion provides no backtest or evidence that MACD on any of these inputs predicts profitable trades.
Tags
Full text
# What is more appropriate: the EMA of the option price or the EMA of the underlying? # What is more appropriate: the EMA of the option price or the EMA of the underlying? I'm progressing, all too slowly, on a site that aims to show real-time numbers for options that are listed on the CBOE. Most of the instantaneous numbers are all set. Now I'm going to pay attention to some of the trends in those numbers: option price, volume, put-call ratio, open interest, and similar. One of the sets of values I want to show for each option is MACD of price (others, too, but my question is about price). EDIT: @Tal Fishman -- I want to show MACD because I understand that convergence/divergence of fast and slow EMA is helpful to a guy trying to decide whether or not to buy/sell an option. Am I quite wrong? SHould I abandon MACD and show something else? If so, what is that something else? I wonder how I should calculate the EMAs that contribute to the MACD for option price: Is it more useful, appropriate, and usual to calculate EMA based on the price of the underlying? Or on the price of the option itself? If the question betrays my naïvté, well, so much the better: I welcome comments that tell me how I'm wrong wrong wrong! Thanks so much! The site is http://www.sellmycalls.com/cgi-bin/chain ## Answer by chrisaycock (score 4) https://quant.stackexchange.com/a/1957 Smoothing the underlying's price movements would decrease the volatility of the underlying, which in turn would lower the computed Black-Scholes option premium. This probably isn't the effect you want. (I also don't get the impression you're displaying fair value anyway, just the market price.) So the EMA must be applied to the option's price. ## Answer by Tal Fishman (score 3) https://quant.stackexchange.com/a/1962 I think concepts like MACD should typically be applied to underlying price rather than the option price. Option returns are not what Meucci calls an "invariant," and concepts such as moving averages need to be applied to (sums of) invariants. What I meant in my comment, though, is that if your users will be using options to make volatility bets, then a MACD of implied volatility could make sense, but I've never seen this.
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.