Another Look at Negatively Correlated Strategies

C2
1

STRATEGIES NEGATIVELY CORRELATED WITH THE S&P 500

On the weekend, I highlighted a screen of 12 strategies that were negatively correlated with the S&P500:

https://forums.collective2.com/t/is-it-time-to-consider-some-strategies-with-negative-correlations-to-the-s-p-500?source_topic_id=11365

If you shifted some of your assets to any of these Monday morning, you probably did a lot better than leaving it with the median strategy here.

I just checked and none of the 12 negatively correlated strategies on my weekend list collapsed or increased its maximum DD today. Indeed, they averaged a slight gain this month (as one would expect for a strategy that is legitimately negatively correlated with the S&P 500).

February Returns

  1. 100 Percent Win: 8.3%
  2. Bitcoin Related Aggr: 5.5%
  3. AI Algorthims 13: 3.5%
  4. Bitcoin Related Trader: 3.1%
  5. forexland: 2.7%
  6. OPN W888: 1%
  7. Gewinn Futures Aggressiv: 0.4%
  8. VIX community: 0%
  9. Payoff Matrix: -0.3%
  10. CommodityFutures: -0.8%
  11. TechSurfer Aggressive -0.9%
  12. The Wizard EUR/USD Forex: -2.5%

Given the 99.4% win rate for the strategy “100 Percent Win,” and the seeming Martingale trading practices employed, cautious buyers should probably be wary of that one, since such systems seem to have a greater tendency to collapse at C2. (If we hadn’t just had a huge drop, I probably wouldn’t have included the last sentence, since I have 2 strategies on the list. Of course, 100 Percent Win’s record so far is excellent, and this time maybe it’s different.)

I should mention that, in putting together my screened list of 12 negatively correlated strategies on this weekend, I had somewhat arbritrarily excluded Futures Proxy from my list. The reasons are raised in this thread:

https://forums.collective2.com/t/futures-proxy-a-proxy-to-vix-arb?source_topic_id=11365

Futures Proxy had a worrisomely high win rate and I wasn’t persuaded that its original bid/ask strategy had a future here. For those who are unfamiliar with Futures Proxy, it debuted in mid-January near the top of the Leaderboard with nearly a +100% return but only a 1.5% DD (as I remember). As of today, it has a 66% DD and an overall -11.6% overall return. It lost 51% in Feb., virtually all of it coming in the last hour of trading on Monday. [Tuesday morning update: Futures Proxy came roaring back this morning and is now up substantially for Feb.]

COMPARISON TO THE TOP 7 STRATEGIES BY C2 SCORE:

Contrast the recent performance of the negatively correlated strategies with the top 7 strategies listed by C2 score on the Grid. Two of those 7 strategies collapsed today and 2 more saw their max DD jump considerably.

  • MCProTrader went from a 32.5% max DD to a 100%-108% max DD (down 108% for February).
  • U Got Mail went from a 25% max DD to an 87% max DD (down 87% for February).
  • Capitalize on Volatility went from a 29% max DD to a 45% max DD (down 45% for February).
  • Options Cash Machine went from a 23% max DD to a 35% max DD (down 29% for February).
  • Of the remaining three strategies among the top 7 in C2 scores, one is down 16% for Feb., another is down 9.5% for Feb., and one is up 0.4% for Feb.

So of the top 7 strategies by C2 score, only one did as well as any of the 12 negatively correlated strategies on my weekend screen. And almost half of the top 7 scorers lost 45-100% of their value in the last 3 trading days.

++++++++++++++

BTW, my Bitcoin Related Aggr is up about 12% over the last 6 trading days.

2

Seeking feedback: One of the vol strategies I’ve been developing (not launched yet) has a slightly negative correlation to S&P 500, running a relatively high percentage of trading days long vol or cash instead of short vol. It cuts back on the overall returns compared to short-heavy vol fund switching strategies, but manages moderate gains during most S&P 500 market drawdowns, and backtesting CAGR in the 50-60% range with maxDD in the 30-40% range (IRA friendly, vol funds only).

Is this sort of profile attractive to investors? It’s not any sort of hedge per se, but might help some investors diversify their holdings.

3

I’m curious what you use to go long volatility: UVXY, VXX, or their options?

The problem I had when backtesting going long volatility is that it appeared to work with optimal fitting, but with the price deterioration working against you, I had my doubts that going long VXX or UVXY would work consistently in practice.

But if your model is good enough . . . .

4

Use VXX typically, strong desire to make it easy to trade, liquid, IRA-friendly. I trade vol strats in my IRA. My more aggressive vol strats have been long vol only ~4-7% of trading days over the backtest period, and long vol + cash days around 15%, but this strat with negative correlation is more like 10-20% long vol days and 30-40% cash days, so roughly half are short vol.

5

Got it. Yes, VXX is the most liquid–esp. now.