Should You Hedge or Diversify?

Thank you @Jessefor another great article.

How about trying to combine two approaches: hedge but replace SPY with naked puts? If you sell the puts slightly ITM (maybe 1%) every month, you should still keep some nice upside in case SPY rallies, get some premium to pay for the long puts, without sacrificing returns? Kind of a mix between Anchor and Momentum.

On 4/24/2020 at 3:40 PM, Kim said:

Thank you @Jessefor another great article.

How about trying to combine two approaches: hedge but replace SPY with naked puts? If you sell the puts slightly ITM (maybe 1%) every month, you should still keep some nice upside in case SPY rallies, get some premium to pay for the long puts, without sacrificing returns? Kind of a mix between Anchor and Momentum.

Using ORATS again, this is what a combination of a 55 delta short put and 30 delta long put looks like since 2007:

CAGR: 3.5%
Volatility: 6.13%
Max DD: 12.82%
Sharpe: 0.57

Not very encouraging..

11 minutes ago, Kim said:

Not very encouraging..

There may be better parameters to run the long put hedge. If you want me to test any other combinations of entries and exits for the hedge, I'm happy to do so. For example, shorter or longer DTE entries as well as exits. A 365 DTE entry and 45 DTE exit made sense to me for managing the negative theta, but there might be different parameters that work better. Tradeoffs to everything, of course.

What I'm doing for my personal portfolio is the following:

• Selling SPX puts one month out around 1% ITM
• Spending 1% of the portfolio on cheap OTM puts (around 2-3 delta) spread over 6-24 months

During the current meltdown, the long puts provided me a very protection dues to the fact that VIX exploded to 80+. I'm not sure how easy it is to test it..

What about hedging with VIX calls.

There are some materials on this topic

https://investps.com/wp-content/uploads/2016/09/PortfolioDiversificationUsingVIXOptions-Final.pdf

https://www.theoptionsguide.com/portfolio-hedging-using-vix-calls.aspx

https://seekingalpha.com/article/613051-buying-vix-calls-as-a-portfolio-hedge

https://steadyoptions.com/articles/hedge-portfolio-vix-options/

Also https://www.soa.org/globalassets/assets/library/newsletters/risk-management-newsletter/2015/august/jrm-2015-iss33-lerner-metli.pdf

They say "We find that if your view is for a large and systemic selloff—greater than what is priced by the market— then a volatility-based hedge can potentially make more sense. For smaller and more-fundamentally driven selloffs, directional hedges via equity options may be the more appropriate choice."

@WayneCthanks for the links, great resources.

I actually looked into VIX calls and compared them to far OTM SPY puts. I looked at the approach of 50 cents trader. Take a look:

https://www.cnbc.com/2018/02/12/the-mysterious-trader-nicknamed-50-cent-made-200-million-last-week-as-the-market-blew-up.html
https://www.reuters.com/article/us-usa-stocks-volatility/big-volatility-options-trade-points-to-mystery-investor-50-cent-idUSKBN1ZS0HW

The strategy involves buying next month OTM VIX calls around 50 cents. 

It paid of very well in the last crash, but the crash was very sharp and quick, and VIX reached 80+ in a matter of few weeks. In case of more gradual decline, the hedge would be less effective, and after the initial decline, the following months VIX futures will become very expensive and not economical for hedging.

Of course you can spread the hedge over few months, but the following months become too expensive. For example, under "calm" markets (VIX in low teens) you can buy the following month 25 calls for 50 cents. But if you go 6-9 months out, then to spend only 50 cents you will need to go up to 40-45 strike.

But maybe some mix of SPY puts and VIX calls can make sense.

Going back to the article, if we take the current crash, we can see that hedging (as represented by our Anchor strategy) worked better than diversification (as represented by Steady Momentum strategy), at least so far. Both strategies reduced portfolio volatility, but Anchor was actually up at some point in March when SPY was down 30%+. @Jesse would you like to comment?

On 5/2/2020 at 7:19 AM, Kim said:

@WayneCthanks for the links, great resources.

I actually looked into VIX calls and compared them to far OTM SPY puts. I looked at the approach of 50 cents trader. Take a look:

https://www.cnbc.com/2018/02/12/the-mysterious-trader-nicknamed-50-cent-made-200-million-last-week-as-the-market-blew-up.html
https://www.reuters.com/article/us-usa-stocks-volatility/big-volatility-options-trade-points-to-mystery-investor-50-cent-idUSKBN1ZS0HW

The strategy involves buying next month OTM VIX calls around 50 cents. 

It paid of very well in the last crash, but the crash was very sharp and quick, and VIX reached 80+ in a matter of few weeks. In case of more gradual decline, the hedge would be less effective, and after the initial decline, the following months VIX futures will become very expensive and not economical for hedging.

Of course you can spread the hedge over few months, but the following months become too expensive. For example, under "calm" markets (VIX in low teens) you can buy the following month 25 calls for 50 cents. But if you go 6-9 months out, then to spend only 50 cents you will need to go up to 40-45 strike.

But maybe some mix of SPY puts and VIX calls can make sense.

Going back to the article, if we take the current crash, we can see that hedging (as represented by our Anchor strategy) worked better than diversification (as represented by Steady Momentum strategy), at least so far. Both strategies reduced portfolio volatility, but Anchor was actually up at some point in March when SPY was down 30%+. @Jesse would you like to comment?

I'm waiting for a response from Matt at ORATS about a question related to the backtesting output...I've found a discrepancy that doesn't make sense to me, and might be causing the backtest to be displaying results that are worse than they should based on if you enter the backtest as a "married put" strategy vs. a "protective put" strategy. I backtested based on "married put", but if I change it to "protective put", the results are much better for the hedged strategies. If that turns out to be true, I'll need to update the article with corrected numbers, but until I hear back from Matt I'm not sure so we'll have to standby.

On 5/4/2020 at 10:17 AM, Jesse said:

 

I'm waiting for a response from Matt at ORATS about a question related to the backtesting output...I've found a discrepancy that doesn't make sense to me, and might be causing the backtest to be displaying results that are worse than they should based on if you enter the backtest as a "married put" strategy vs. a "protective put" strategy. I backtested based on "married put", but if I change it to "protective put", the results are much better for the hedged strategies. If that turns out to be true, I'll need to update the article with corrected numbers, but until I hear back from Matt I'm not sure so we'll have to standby.

I've updated the data in the chart based on discussions with ORATS to make sure the backtest is being simulated properly. The hedged portfolios now looks better than before, although still less attractive on a risk-adjusted basis relative to stock/bond portfolios that have been scaled to have the same volatility as the hedged portfolios. The hedged portfolios also don't factor in the cash outlay required to buy and roll the hedge, so this essentially simulates buying the puts interest free on margin. If that were properly accounted for, by either deducting applicable margin interest, or if not using margin reducing the quantity of SPY, results would be slightly worse than shown.