2

Imagine someone bought 100K SPY as a long term investment. Now he wants to hedge against the downside risk of 10% or more. He is considering the following options:

  1. Buy UVXY which is a 1.5X VIX ETF inversely correlated with SPY. If one buys 5K, in case of a 10% SPY crash, one would make ~50% or 2.5K (not sure how to calculate). Pros: No time decay. Cons: Price can drop very fast due to quick SPY upward movement. No exposure to change in volatility. Unpredictable upside.

  2. Buy inverse 3x SPY ETF SPXL. If one buys 5K, in case of a 10% SPY crash, one would make ~30% or 1.5K. Pros: No time decay. Uniform 3x daily return compared to SPY Cons: Price can drop very fast due to quick SPY upward movement. No exposure to change in volatility. Limited upside.

  3. Buy monthly or longer PUT options. If one invests 420 per month or ~5K/year, 10% drop would give 10X or more return Pros: One gets exposure to both change in price and volatility. Cons: OTM put options can lose value quickly due to time decay. ITM calls can be expensive depending on volatility.

Now how can one find out which option is the best and what percentage of investment to put there? Is there any way to backtest different strategies?

pappu
  • 139
  • 3
  • 1
    Although of course the future will not be like the past, for a problem like this backtesting (on real or simulated data) is indeed the way to come up with a first evaluation of the costs and benefits of these strategies. Without this it is just armchair theorizing. – nbbo2 Mar 11 '21 at 20:40
  • Yes I was trying to do that in python, not sure where to get the option data though. – pappu Mar 11 '21 at 20:42

1 Answers1

2

re 1 and 2: You might want to Google something like 'volatility decay' or 'volatility drag' for leveraged ETFs.

re 3: a quick look at http://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/histretSP.html suggests the long-term average return for US stocks is ~11.6% pa. You're proposing giving up ~5% pa to hedge. Not clear that's a good idea.

also re 3: if you want to hedge, it's not obvious that simply buying and rolling puts is optimal. At the very least, you'd want to think carefully about expiry, strike, rolling strategy, monetisation and whether or not you delta hedge.

My 2c: given that long-term realised vol in DM equities is ~20%, it seems odd to me that you'd want to hedge a 10% DD. I'd simply reduce the SPY position to a level where you can live with the vol/DD.

user42108
  • 2,252
  • 5
  • 10
  • 1
    Ruling out leveraged ETFs due to volatility drag is total innumerate nonsense. Drag is obviously a continuous function of the amount of leverage you take, so if 1x leverage is better than 2x, why not .5x leverage? 0x? There must be an optimal positive amount of leverage and there's no reason it should be 1x. – Kevin Carlson Mar 11 '21 at 21:39
  • I didn't rule out leveraged ETFs, I simply suggested the OP investigate the vol decay/drag. The OP refers to SPY as a "long term" investment, for which a leveraged ETF might not be an appropriate hedge. E.g. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1404708 – user42108 Mar 11 '21 at 23:26
  • My understanding is that volatility drag is low over a long term for assets with low volatility like SPY. So a portfolio of 3x SPY ETF SPXL and 1.5x VIX ETF UVXY as hedge sounds like a good option. – pappu Mar 13 '21 at 21:23