Hedging trend following - critique this plan?

[amooseman]

Surprised no one has pointed out that long 100 shares SPY + 1 LEAP put is synthetically equivalent to buying 1 LEAP call at the same strike. You can free up a ton of capital by just buying the call. Maybe sell some short term OTM calls to reduce carry.

That's an interesting idea. Thanks for the study Reformed Trader.

Here's another variant of the trend following strategy I'm playing with:
- Each month, short 1 at-the-money to 1% in the money SPY put if it's above the 200 dma
- Hedge with the SPY put leap and sell 1 SPY put 10% below the leap each month to create a diagnol. Same hedge as the initial strategy.

The advantages of the short put vs. long SPY:
1) Margin
2) It better handles one of the disadvantages of trend following - weakness in flat and volatile markets. The short put will earn more premium in volatile markets and will still perform well even if the market is flat vs. long stock.

 

[Reformed Trader]

Yes, using the short put instead of long stock should work a lot better. I have seen a paper about this but haven't had time to read it yet:

Momentum and Covered Calls Almost Everywhere
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2921778

From my brief look through the paper, I would add the caveat that while trend-following is a positive autocorrelation strategy, covered calls/short puts are taking advantage of the volatility risk premium as well as negative autocorrelation, and the VRP effect is probably larger.

The time period chosen for the study also coincides with multiple bear markets (the U.S. market being an exception), so again, the trend-following and short call strategies are going to backtest well relative to the underlying indices.

Interestingly, the conclusion suggests that covered calls had zero correlation with trend-following but that short straddles had a negative correlation (the straddles have no market beta).

Pairing the put trend-following strategy and the diagonal put strategy could potentially create margin calls if the market moves up a lot (forcing some of the short puts to not be completely covered by the LEAPS) and then down quickly.

 

[amooseman]

Yes, using the short put instead of long stock should work a lot better. I have seen a paper about this but haven't had time to read it yet:

Momentum and Covered Calls Almost Everywhere
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2921778

From my brief look through the paper, I would add the caveat that while trend-following is a positive autocorrelation strategy, covered calls/short puts are taking advantage of the volatility risk premium as well as negative autocorrelation, and the VRP effect is probably larger.

The time period chosen for the study also coincides with multiple bear markets (the U.S. market being an exception), so again, the trend-following and short call strategies are going to backtest well relative to the underlying indices.

Interestingly, the conclusion suggests that covered calls had zero correlation with trend-following but that short straddles had a negative correlation (the straddles have no market beta).

Pairing the put trend-following strategy and the diagonal put strategy could potentially create margin calls if the market moves up a lot (forcing some of the short puts to not be completely covered by the LEAPS) and then down quickly.

Awesome study! I had no idea there were so many covered call indexes. I'm glad to combine uncorrelated components into a strategy to further reduce volatility and drawdown.

Of course, as you warn, a far left tail move wipes out most of the benefits. The leaps/diagnol strategy is the best I've come up with so far to cover the far left tail moves. I'm aware of ratio backspreads but have never studied them quite closely. Maybe they would make a better hedge? The diagnol seems very straightforward to me comparatively.

 

[Reformed Trader]

There are different kinds of tail moves, and they can't all be hedged the same way:

1. Large, sustained down moves in the market (2000-2, 2007-9)
2. Large, sustained up moves in the market (2017 and Jan 2018)
3. Sudden down moves in the market (1987)
4. Large implied volatility spikes that aren't accompanied by correspondingly large market moves (Feb 2018)
5. Drawdowns in specific strategies (cheap stocks have underperformed expensive stocks from 2007-2018; momentum had a large drawdown in 2009; trend-following has underperformed from 2009-2018)

My own philosophy is to hedge all of these possibilities separately. IMO, #5 is particularly difficult to hedge. It's possible to get away without hedging them, but that makes leverage very dangerous.