I am a firm believer in Dr. Thompson’s mode of learning the “music” of writing by typing out verbatim writers he admired original copy.
I find this a great way to learn about a new topic of interest, especially in the heady world of high finance. So this morning I’m learning from UBS’s Interest Rate Analysts about the right environment to sell volatility on within the interest curve.
The concept of selling volatility should make intuitive sense because we’ve seen interest rate volatility spike on two separate occasions in 2018, steepening the implied volatility curve. Or in other words, shorter expiry is cheaper than longer expiry, making it cheaper to hedge a short volatility position.
In a hiking cycle selling vol makes sense but how do we balance trade risks
We have highlighted that the recent hiking cycle has been defined by a decline in rate
volatility (Figure 1).
Therefore, we have generally advocated selling interest rate volatility . We have seen this dynamic play out in this hiking cycle as well. We think the key reason selling volatility works is that the Fed hiking removes upside tail risks to inflation/economy. That said, we have noticed more volatility spikes in 2018 than 2017. Specifically this year, we saw a VIX spike in February and Italian sovereign disruption in May (Figure 2).
Looking ahead, we envision trade war-related uncertainty over the next few months. Meanwhile, interest rate volatility curve (for gamma) has steepened (Figure 3). We think this dynamic offers a cheap hedge to selling volatility. Specifically, we recommend selling 1y10y straddle hedged by 3m10y straddle. Next, we show why such a trade setup makes sense based on recent historical experience.
What’s the best way to hedge against a vol spike?
Consistent with intuition, buying shorter expiry options and selling longer expiry
options reduces risks of selling volatility. Specifically, we look into the volatility spikes
(Jan 15-Feb 8) and Italian Sovereign bond led risk-off (May 11-30). 1y10y volatility
jumped 5-15bp over these two episodes. As shown in Figure 4, a short straddle at
1y10y point lost 14bp. To the contrary, a calendar straddle (buying short-dated and
selling long-dated) can reduce this loss by half. Looking ahead, we see an increase in
the probability of trade war escalation and the Fed having to skip its December hike. In
this environment, we think near-term uncertainty can pick up. Thus, we think it makes
sense to sell longer-term volatility but hedge some near-term trade risks by buying
A backtest shows hedged sold vol position is effective when vol curve is steep
Hedges do not come for free. The premium of short-expiry straddle could be expensive,
and roll-down could be painful. But in our case, the hedge is attractive because the
implied volatility curve is getting steep (Figure 3).
Said differently, it means short expiry is cheaper relative to longer expiry, making it cheap to hedge a short vol position. To add evidence to this view, we backtested the performance of 1) short 1y10y straddle vs 2) short 1y10y straddle hedged by long 3m10y straddle (gamma neutral).
On the hedged trade, we only evaluate the performance when 1y Z score of 1m10y-1y10y
volatility slope is greater than one. We look into 1month holding period, and data
ranges from 2013 to now. The results show that the gamma hedged short is
comparable to outright short in both hit ratio and P&L (Figure 5 and Figure 6).
Short 1y10y straddle hedged by long 3m10y straddle
Based on the analysis above, we recommend short 1y10y straddle hedged by long
3m10y straddle. The notional ratio of 1y10y vs 3m10y is 1:0.43, keeping the structure
gamma neutral. The structure has 43bp premium intake, negative vega, and positive
theta (Figure 7).
A scenario analysis matrix is presented in Figure 8. The lower triangle
is corresponding to flatter volatility curve. The flattener loses when 1y10y volatility rises
by more than 5bp (The loss is still smaller than bare short of 1y10y vol straddle).
Although the structure tends to lose more money when volatility curve steepens instead
of flattening, the chances of steepener are slim, in our view, given the current pricing
and macro backdrop.