This blog post is going to be my reminder to myself to really check my risk management and to avoid my biases. This Monday, February 5th, 2018, was the worst return I’ve experienced in my portfolio, although I’ve managed to recover a bit, that was based on pure luck and I should never be caught in this situation again.
Ultimately, implied volatility (the VIX index) spiked 100%+ on Monday for reasons I’m speculating to be a cascading effect of volatility algorithms and funds getting spooked by the initial 30% spike on Friday. Whatever the reason, all my research, back-tests, and codes were thrown out the window and ignored by me. I was supposed to be systematically putting on my short SPX put positions so that I would avoid exactly this scenario I got caught in: overleveraged on SPX puts and having to buy them back at a significant loss.
I’ve been trying to think over all things I’ve done wrong because this strategy – even though I’m still left in shock – is not one I will discontinue. In fact, when implied volatility spikes like this, it’s actually when the most money can be made because of the irrationality of the markets in panic mode. At moments, SPX put options for 20% drawdowns for 1-day expiry were selling for $10+ a contract. These were actual goldmines given that full day market circuit breakers were triggered at 20% loss on the index meaning that these options would be almost certain to expire worthless. However, I was in panic mode and lost my ability to think clearly and was completely paralyzed from taking advantage of this opportunity.
I’m going to list out, to the best of my recollection, what happened in my mind and what I did.
- Wednesday, January 31st: I ran my usual SPX Implied VaR to get the 0.05% loss and compared it to the worst case scenario for the lower VIX level of 14 and decided that selling a February 5th 2680 SPX put option was safe. In hindsight, I made the mistake of not using the open-to-close check and just used the close-to-close check given that I was putting the contract in mid-day.
- Things were going well on Thursday with the market pretty flat and VIX keeping steady; I was getting paid the theta on the contract so I didn’t bother.
- Friday, February 2nd: I decided to go snowboarding, which was fine, however, I completely ignored the markets and didn’t check until the end of the day where I realized that VIX had risen 30%+ from the day before and that my positions were down, in what I thought at the time was a lot. At this time, I should’ve re-ran my checks because the volatility regime has now changed and we were in mid-level VIX territory. I completely ignored Bayesian probability. Had I done so, there may have been a chance that I would’ve closed my positions out as soon as I could (which would’ve been the following Monday morning).
- Monday, February 5th: This was the day of reckoning and one where I completely dropped the ball on execution. During the morning, there was a slight rally in the S&P 500 and I could’ve closed all my positions with a net positive – I had three hours to do so, but I didn’t. VIX had climbed to 18 by this point, which should’ve been another flag had I chosen to ignore the check on Friday. By 2:00 pm, things were going quite south and I was down the same amount as I was on Friday. My sunk cost fallacy bias kicked in along with the panic and I thought I could hold it, but when things turned really ugly and liquidity dried up in the options books, I closed my position at a substantial loss. To make matters worse, I decided to initiate a 2600 SPX put in the morning for February 7th expiry and then during the afternoon, when I thought the worst was behind me, I decided to initiate another 2500 SPX for February 7th as well. This all happened before 3:00 pm where the majority of the selloff took place. Before I knew it, I was down multiples more before the close and I didn’t close anything. I managed to slightly offset some of my losses by putting on another position, this time at 2000 SPX for almost $5, but I couldn’t put on more because my risk-reversals already held in the book along with the other two contracts had pushed my margin to the limit.
- In hindsight, I should’ve taken heed to the warnings in the morning as there was ample time and closed everything. Then, I should’ve waited to just see if the storm was going to settle before initiating any more positions and watched the options chains very closely and take advantage of the liquidity by selling extremely overpriced and near impossible strikes. Doing this would’ve probably netted me almost 5% gain through the entire turmoil.
- Tuesday, February 6th: Luckily – and this was extreme luck – the markets didn’t continue their slide this day, but I didn’t want to risk anything as my 2600 and 2500 strikes were both getting tested so I closed them out for even more losses. I did manage to get a 2200 strike for another $5, which would’ve meant an 18% decline in the S&P in 2 days, which was highly unlikely. I also closed all the puts in my risk-reversals as they were eating margin room. I was lucky to come out of this without losing more than 3%, however, all my effort and research should have allowed me to avoid this very scenario.
Overall, the short SPX strategy is still highly profitable and I should not shy away from it even though I’ve been burned. It wasn’t the strategy that was poor, it was my execution and risk management. I was overleveraged, panicked, and ignored red flags and warnings. I was also too complacent with the markets, which led me to ignore actively managing my positions.
In the background, the ETFs XIV and SVXY were both completely wiped out from this turmoil and this was another reminder to be more systematic with, not only opening my positions, but exiting them early even at a loss. My long XIV signal had exited the ETF long before this debacle and completely avoided the losses.
So, for my sanity and future trades, I’m going to list out how I’m going to systematically sell SPX to avoid this again, and hopefully be on the opposite of the trades should this happen again.
- During low VIX periods, sell at maximum 3 days to expiry and watch the implied volatility level. If the VIX spikes more than 10%, re-evaluate your position given the current data and see what the worst-case scenario is. Close positions if it breaches. Do not put on more than 10% of your margin room during these periods.
- Instead of SPX options, use ES options, which are half the size, but can be closed overnight offering more flexibility in risk management. In addition, because they are not cash settled, you can take the futures assignment, which should only be done if you strongly believe the market is not entering bear territory, e.g., macro fundamentals are still very healthy. As well, you can sell calls to offset some losses and reduce your cost-basis.
- Watch the VIX term structure, since, selling puts is essentially shorting volatility, a backwardation should be a red-flag telling you to considerably reduce your exposures, my guideline is to reduce 50% of your current exposure.
- If VIX is approaching 20 from below and you have more than 1 day until expiry, close out the positions early and eat the loss. This may hurt as well sometimes if there is a reversal and your positions would’ve remained safe, but that’s just hindsight bias and you’d rather take the loss you know than risk the one you don’t.
- If VIX remains high, you should be sitting almost completely on cash. Do not trade in the morning even if things look like they’re reversing. Wait until near the market close, around 3:45 pm to 3:55 pm and run your options VaR script and worst-case scenarios. If there is an opportunity where you see severe lack of liquidity and irrational prices, e.g., a -20%+ contract with 1-day to expiry is selling for $5+, sell up to 50% or more of your margin for those.
- Otherwise, when in a high VIX environment, sell only 1-day expiries very far out the money – specifically below the historical VaR – and only sell a couple contracts. The IV should still make these contracts very rich in premiums so you only need to sell a few.