Turning the study on it's head and buying calls is pretty straight forward - flip the signs and take the other side of the bid/ask spread. You win about 27% of the time, and make *almost* as much as if you just bought the stock.
If instead of buying 10k shares in 2019, you instead bought 25 delta calls every month, you'd end with $32M vs. $36M. That assumes you consistently buy at the same ratio as the overlay, probably not practical from an options buyers perspective.
While you could get more efficient with capital, sizing long buys should be conservative. The calls cost roughly 3-8% of your bankroll at any given time.
I was not sure how the bid-ask spread might impact things.
Pretty sure that the delta in your experiment was 25d. I went to https://stendec.io/ctb/bscalc_simple.html and ran the following: S=400, X=450, DTE = 25 (T=0.0685), vol=60%, r=b=0. That gave me 25d and N(d2) is .204. so risk neutral probability of exceeding strike is calculated as 20.4%.
Yup, 27% is definitely higher than that. Stock has risen quite a bit, which I think makes this interesting to compare to the attribution work Kris did. If you delta hedged the calls, it isolates the vol pricing whereas the path of TSLA really impacts the long/short call.
The only reason I bring up bid/ask is that from a raw PnL perpective long calls isn't the inverse of short calls - you're giving up spread in both cases.
Thanks. Yes, I'm glad you talked about the bid-ask spread. That was exactly why i was not sure that the inverse pnl would be positive (and certainly not as good as it was).
All very good and interesting questions!
Turning the study on it's head and buying calls is pretty straight forward - flip the signs and take the other side of the bid/ask spread. You win about 27% of the time, and make *almost* as much as if you just bought the stock.
If instead of buying 10k shares in 2019, you instead bought 25 delta calls every month, you'd end with $32M vs. $36M. That assumes you consistently buy at the same ratio as the overlay, probably not practical from an options buyers perspective.
While you could get more efficient with capital, sizing long buys should be conservative. The calls cost roughly 3-8% of your bankroll at any given time.
Thanks, Mark.
I was not sure how the bid-ask spread might impact things.
Pretty sure that the delta in your experiment was 25d. I went to https://stendec.io/ctb/bscalc_simple.html and ran the following: S=400, X=450, DTE = 25 (T=0.0685), vol=60%, r=b=0. That gave me 25d and N(d2) is .204. so risk neutral probability of exceeding strike is calculated as 20.4%.
27% win rate is FAR higher.
Does that make sense?
Yup, 27% is definitely higher than that. Stock has risen quite a bit, which I think makes this interesting to compare to the attribution work Kris did. If you delta hedged the calls, it isolates the vol pricing whereas the path of TSLA really impacts the long/short call.
The only reason I bring up bid/ask is that from a raw PnL perpective long calls isn't the inverse of short calls - you're giving up spread in both cases.
Thanks. Yes, I'm glad you talked about the bid-ask spread. That was exactly why i was not sure that the inverse pnl would be positive (and certainly not as good as it was).