How To Design A Robust Trading Algorithm

For Systematic Premium Buying

As market dynamics evolve, new opportunities pop up and require new approaches to lock in profits.

The first step in the process is to identify the market factor that recently changed. One change that occurred recently is the complete erasure of the risk premium in major index implied volatility. Yes, the decline in index volatility has been going on for decades but recently, the implied volatility levels have fallen so low that it raises the possibility that systematic premium buying might actually be a winning strategy. This is in stark contrast to years gone by where traders used to sell index volatility day in and day out, earning consistent (albeit high risk) returns from selling premium.

Zero Days To Expiration Options offer an idea way to test out systematic premium buying strategies. Because they expire in one day, they offer the lowest cost of entry because most of the time premium is taken out of options up and down the chain. More importantly, the sheer number of trades provides more data points for test trades and thus, quicker feedback on algorithm strengths and weaknesses.

As we design and deploy the algorithm, we need to answer of few questions:

• What strategy will we use? Long Straddles? Long Verticals? Short Butterfly Spreads?

• What additional parameters will we need for trade entry? For trade exit?

In my experience, the fewer parameters the model contains the better. This reduces the risk of over fitting the trading model to past data. For every extra parameter that we add to the model, we introduce another factor that has to be correct if we want to make money. It is better to have something work out well if only a few stars have to line up. If we look back at crowd sourced backtest platforms like Quantopian, we see that 100,000 active users produced no winning strategies at all on any of the strategies they developed, even with massive financial backing from very big names in the hedge fund industry.

The big drawback of using only a few parameters is that the parameters chosen have to be correct.

So how do we answer the critical questions concerning what is the best strategy to use, and the correct parameters to apply?

Back in my consulting days, one of the top engineers I met in my life told me:

Good programmers write great code — in a short period of time. Great programmers write tools.

If you are a paid subscriber to this Substack, post the word CODE in the comments and I'll send you a set of C++ software tools you can use to explore the world of systematic premium buying.

I will be developing this concept over the next few Substack posts and show you the results as they come in.

NOTE TO PAID SUBSCRIBERS:

The Air Raid Signal will be active tomorrow because the index futures are down sharply overnight. This signal doesn't get triggered often, if it goes off I will send out an alert via Substack.

DISCLAIMER: All content on he Nuclear Option Substack is for Education and Information Purposes Only. It is not a recommendation or solicitation to buy or sell any security. Before trading, consult your Professional Financial Advisor and read the booklet Characteristics and Risks of Standardized Options Contracts, published by the Options Clearing Corporation.