How could I calculate the probability of getting wiped out?

Question

Last year I ran a mildly stupid/naive options strategy.

It involved picking 10-15 random companies, and then setting up put credit spreads on each one, expiring 1-2 weeks out. I prioritized high volume, blue-chip underlyings.

I would exit the trade at either 60% profit, or 100% loss. If the underlying was 1 DTE, then I would let it ride to get that sweet, sweet last-day theta decay.

On average, I would get a 10%-20% return on a weekly basis. My worst week I was down 40%.

I ended 2019 up 3000% from where I started.

Both fortunately and unfortunately, I had a medical procedure that I needed to fund in 2020. I took all of my money out of the market to fund it. If I had kept this strategy going, I would have most certainly been wiped out given the recent market turmoil.

My question is: is there a way to calculate the probability of getting decimated by a strategy just like this? If you knew all of the intrinsic and extrinsic value of your options -- the Greeks, etc. -- is there a way to get a running probability of decimation in a spreadsheet or algorithm?

This question is more of a fun exercise in modeling, rather than something that I would actually employ.

I'll likely use a strategy like this when we get back into a strong bull market, I just figured this might be someway to occupy myself while I sit on the sidelines.

Answer 1

The answer is yes. There is a way to estimate this and doing those sorts of estimates for portfolios that may include options positions is a big part of what programmers/traders/risk-managers do at banks, trading firms and financial institutions.

You seem interested, but understand that this may be over your head. If it is, then I'd suggest maybe some online courses that indicate that upon completion you'll be capable of calculating VaR for a portfolio. That's not exactly what you want, but it's close enough that if you can do that then you should be able to do what I describe below...

First, you need a covariance matrix for your stocks using say the last 3 months of data. You should "shrink" the covariance matrix to bias it to be "worse" for your portfolio because correlations tend towards 1 in a crash (look up shrinking covariance matrix). If the positions are relatively short time to expiry and limited loss and gain you can then run simulations and look at the outcomes to price your options at expiry and see what the distribution of returns looks like. Meaning, you don't have to do options maths to price them prior to expiry.

As far as implementation: Python is cool and you wouldn't regret learning it. This is a lot though based on what I'm hearing from you, but you did say you're laid up, so... have fun! And check out quant stack exchange.

Answer 2

I'd say that you ran a mildly smart option strategy because:

• You weren't chasing naked premium (the long leg prevented disaster)

• Your time frame was very short (two weeks) so theta decay was high and accelerating

• You had exit rules

Apart from the above, the reason that you succeeded was that for the most part, 2019 was an up trending year. Bullish credit spreads do well in that kind of environment. The secondary reason is that you picked quality underlyings that cooperated (Blue Chips tend to track the market) and picked few dogs that went into the crapper. Otherwise, you wouldn't have done as well as you did.

To survive a tail risk wipe out like this year would require owning negative delta:

• You could achieve that with a mix of bullish and bearish positions.
• You could ratio your spreads, buying slightly more long puts than short.
• You could defend positions, rolling the long puts down and with each roll, slightly increase the number of long puts, turning the position into a backspread. This did quite nicely in March though the massive increase in implied volatility made this harder later in the month.
• You could flip from bullish to bearish verticals due to market conditions but this one is far easier said than done given the speed of March's descent.
• You could buy some OTM index puts as a global hedge.

Regarding the last one, due to the market run up, for the past few years I have purchased one year or longer SPY LEAP verticals with the long leg 10% OTM and the short leg 20% OTM as a portfolio hedge. Cost has been approximately 1.5% or so of the principal. With a few leg out and ins of the short leg during the year, the total cost was reduced to under 1%. Despite this, it was a total waste of money in 2018 and 2019.

Their purpose was to hedge a chunk of my long equity exposure. Toward the end of the year the short leg tends to be near worthless and I cover it, ending up with decayed long puts. Several months before they expire (before the accelerated last months of theta decay), I roll them into the next year's new protective vertical. Needless to say, those long puts paid off quite nicely this year, massively reducing my equity losses.

Per the answer by RWP, you could delve deep into statistical analysis and derive a precise probability based on some period of historical data. It's just my opinion but I think that's a bit of a wild goose chase because it's based on data from previous months and a tail risk event like March doesn't care about your probability number. Your survival will be based on the choice of strategy(ies) as I listed above as well as your ability to manage and adjust your positions or shut them down to cut your losses (active risk management).

Lastly, some suggest that the delta of an option provides a loose approximation of the probability that the option will finish ITM. Probability Of Touch is about double that. I'm not too keen on this because implied volatility affects delta, more so when OTM so such a probability number can fluctuate significantly. Ball park number? Yes. Accurate. Methinks not.