The Saturday Spread: Using the Markov Property to Find Mispriced Opportunities (PANW, NTES, DKS)

The Saturday Spread: Using the Markov Property to Find Mispriced Opportunities (PANW, NTES, DKS)
The Saturday Spread: Using the Markov Property to Find Mispriced Opportunities (PANW, NTES, DKS)

It is inevitable that on any given day, Wall Street will be mispricing the option premium on a publicly traded security. Specifically, the standard Black-Scholes model effectively states the following for debit-based transactions: Assuming the stock moves randomly with constant volatility and no memory, the fair price of a call option is the expected discounted benefit of holding the stock above the strike price at expiration.

As such, the model provides a clean template as a reference point but without much contextual support. Before I am inundated with emails from angry pedants willing to defend the honor of Black-Scholes, let’s really consider the trifecta of why I made the above statement. We know that:

  1. Stock market movements are not random (as we observe autocorrelation and clustered behavior).

  2. Volatility is not constant (as it typically expands and contracts depending on the underlying catalysts).

  3. Actions have memory (since what happened before affects what may happen later).

In fact, the last point about market memory is one of the philosophical foundations of the Markov property. Under this framework, the future state of a system is determined solely by its current state. In other words, according to Markovian reasoning, the fulcrum of transitional logic centers on the immediate behavioral state. According to Black-Scholes, behavioral states are not considered, neither in the immediate framework nor in the deep past.

To be clear, this lack of calculation does not make Wall Street’s standard pricing mechanism incorrect, but it does make the projections generated potentially suboptimal. This is because according to Black-Scholes, since the state context is not considered, risk is largely defined in proportion to the distance of the location. That’s like saying a 3-pointer is harder to make than a layup, which is usually a reasonable statement.

However, in real game conditions, the path to the layup could be heavily defended. In that case, the open player outside the arc may have the easier shot, even though the distance is greater. That’s basically the Markov property. It is a second-order analysis that derives probabilities from context rather than model assumption.

Let’s get to work. Palo Alto Networks (PANW) has a spot price of $187.68 at the time of writing. According to the Black-Scholes-based expected movement calculator, for the options chain expiring on February 20, PANW stock would be expected to gain between $171.31 and $204.01. Since this range represents a perfectly symmetrical high-low spread of 8.71%, one can see the potentially suboptimal nature of the price dispersion.

Source link