The ES/SPY on a day-to -day basis is actually less volatile than a 10-year bond, but TLT and IEF have much shallower skews and IV than SPY. But look at the huge trading range for TLT since 2009. It has much bigger swings than SPY, and it's not uncommon for TLT to fall 5-8% in a month whereas that is very rare for SPY. I would seem obvious that one should sell index volatility and go long bond volatility. This seem to represent some sort of fundamental mispricing or complacency in the treasury bond option market, or unless I'm missing something here.
treasury bonds more volatile than stocks but have smaller IVs
- Thread starter tradelosses
- Start date
backtesting shows good returns buying TLT 90-day put spread when it's near 52 week highs. the first leg is 8% OTM and the second is 11% , while selling the ATM strangle to help cover the premium. Once the target is met, one should go long again.
Bonds have a much greater tendency to mean revert than index funds but don't have the huge IV skew that make buying puts unprofitable
Bonds have a much greater tendency to mean revert than index funds but don't have the huge IV skew that make buying puts unprofitable
Last edited:
That's a complicated way to express a simple directional view. Why not just short TLT, or buy some puts?
I think it might be wise to look at the period before 2009 when exploring this...
It will be a small-number-of-observations problem no matter how far back he looks in the data.
either it falls off a cliff or it doesn't. if it does like in 2009 or 2013, you make a lot off $ from the bear spread (more than you would from simply shorting it). If it's flat , you break even from the premium of the strangle . the low IVs for in money options make this possible because the ATM IV is as high for TLT as it is for SPY but the ITM IV is much lower
Doesn't "shallow" skew argue against doing the trade in put spread format? And selling strangles?
shallow skew benefits spread buyers
Surely it would depend on the strikes, no? I just priced the 3m TLT put spread using the "steeper" SPY skew and I see it costing more with the "shallow" TLT skew. To be fair, this is with the upper strike being ATM or close. You're correct for strikes further OTM.