Tag: butterfly

Using Momentum to Short Butterflies

Turning Momentum’s primary weakness into strength

The biggest problem with any trend-following strategy is sudden reversals in trend. This fits nicely into the biggest problem with selling butterflies – that the underlying stays where it is. One thing that momentum stocks guarantee is that they will move big.

Our earlier attempts at developing a mechanical short-call butterfly strategy for the NIFTY failed because there would be months when the NIFTY would just end flat for expire very close to where it started after bouncing all over the place. Let’s see if using single-stock options overcomes this problem.

A quick back-test

In order to hedge our bets, we consider only the top 3 and the bottom 3 stocks in our momentum universe. And then we apply a liquidity filter to ensure that we can actually carry out the trades. Trades are entered 10 days after expiry and held till expiry. No stop-loss is applied.

Top
top mom butterfly
Bottom
bottom mom butterfly
Combined
total mom butterfly

Conclusion

Using momentum to drive a short-call butterfly strategy holds promise. We have created a Theme for it – Shattered – to help traders access this strategy.

Short Butterflies with a Delayed Fuse

Delayed Entry

In our previous post, we saw how a mechanical expiry-to-expiry short-call NIFTY butterfly with a stop-loss significantly outperformed a non-stop-loss strategy. However, the odds that a stop-loss would be triggered was almost 50%.

In our introductory post on butterflies, we had noted how the strategy P&L gets “pulled” as we get closer to expiry. This implies that during the first half of the trade, the position sits idle. However, the underlying NIFTY index could have already made its move out of the wings and could be on its way back to the center. To avoid this, we can delay entry until, say, 10 days have passed since expiry.

Returns with stop-loss and delayed entry

butterfly.returns.SL.10d.NIFTY

short call nifty butterfly 10day delay
The odds of a stop-loss being hit remained the same. Even though the total return improved, when you look at the yearly rollups, the hit-rates are sub-par.

Conclusion

Although on the face of it, selling butterflies on the NIFTY sounds attractive – NIFTY is a volatile index and a rising NIFTY will make the moves required for a profitable trade lower – a mechanical backtest did not lead to a strategy that could be applied on an on-going basis.

Selling NIFTY Butterflies with Stop Loss

Stop loss

We saw earlier how infrequent but large losses in the short-call butterfly strategy can wipe out all the past profits earned. To figure out a stop-loss mechanism, we plotted individual expiry-to-expiry butterflies to find out how each one of them behaved. This data is available in the appendix below.

The stop-loss is set at 10 points (x lot-size) for the butterfly. But this kicks in only 10 days before the contracts expire (see how it gets “pulled” here.)

Returns comparison

Here’s the month-over-month returns of a mechanical short-call butterfly strategy with stop-loss:
butterfly.returns.SL.NIFTY

Here’s the month-over-month returns of a mechanical short-call butterfly strategy without stop-loss:
butterfly.returns.noSL.NIFTY

Because of the stop-loss, the frequency of losses increases but their magnitude decreases. However, stop-losses were hit about half the time.

With stop-loss:
butterfly returns with stop loss
Without stop-loss:
butterfly returns without stop loss

Summary

Applying a stop-loss to a mechanical expiry-to-expiry short call butterfly strategy on the NIFTY seems to enhance returns by cutting out infrequent but large losses. However, the frequency of wins seems low given that only two out of 5 years resulted in net profit. If you did not use a stop-loss, you would have larger gains and would have had more positive years. However, losses in 2010 and 2012 were enough to wipe out all profits and more.

Appendix

https://www.scribd.com/doc/271141815/Nifty-Butterfly

BANKNIFTY Butterflies

Introduction

So far, we have focused on the NIFTY for selling butterflies (Part I, II, III.) How would this look on the BANKNIFTY?

CNX BANK Index Returns

First, let’s have a look at the 30-day rolling returns of the CNX BANK Index, from 2010 to the present, the whole population:

day-30-rolling.BANKNIFTY.2010-2015
Median: 1.79%

2014-present:
day-30-rolling.BANKNIFTY.2014-2015
Median: 2.21%

Beginning of 2015-present:
day-30-rolling.BANKNIFTY.2015-2015
Median: flat

Currently, the index is around 18700. That means a 100-point move translates to 100/18700 ~0.5%

Expiry-to-Expiry Back-test

If we do the same back-test we did to the NIFTY, this is what we find:

butterfly.returns.BANKNIFTY

Conclusion

As with the NIFTY, the trade makes money if you know how to cut your losses. However, when the trade is live, how do we know what the future volatility of the underlying is going to look like? Without risk-management, a short-call butterfly strategy will encounter out-sized losses that wipe out all prior incremental gains.

A Simple NIFTY Short Call Butterfly Back-Test

Monthly moves

We saw that the NIFTY’s median move over a 30-day period is over 1% and that is all that is required to make a short-call butterfly strategy profitable. Let us now do a quick back-test to check if it is indeed the case.

Back-test

Here is the short-call butterfly back-test from 2010 through now. You basically sell the closest expiry butterfly at each expiry (click to embiggen):

butterfly.returns.NIFTY

Summary

While it is true that a 1% move in the NIFTY results in a profitable trade, there are instances where the NIFTY doesn’t move +/-1%. When the NIFTY ends flat, you end up losing all your prior profits. Understanding what drives NIFTY to move is key to managing your risk while running this strategy.