So if (big assumption) annual log-returns are iid normal with known mean m and variance v, then the long-term return after T years is also normal (call it X) with mean mT and variance vT. The capped ETF long-term return is sum of truncated iid normals with support (a,b) which not sure what it is (call it Y) but can be parametrized in terms of m,v,T,a and b. I wonder how E[X-Y] changes depending on the truncation bounds? For instance if m=5% and the bounds 0% and 10% are symmetrical around the expected annual return, then both index ETF and buffer ETF should return the same long-term. In practice, m~7-10% so the bounds are asymmetrical, which erodes the long-term return for the buffer ETF.
I think you're asking a good question -- whether the buffered ETF is a good deal depends on the bounds, so it would be interesting to see for what values it's a good deal or not. But you don't need to assume that the distribution of returns is normal or iid. You can use whatever model of returns you want, and compute the results by simulation. That way you don't get just E[X-Y], you can get the whole distribution.
Good article. The idea of rip off ETF coming from option trading, buy call (K=current price) and sell put (K=current price*1.1) which will secure profit from 0 to 10%. The good thing about this trade is that it is partially self financing, meaning you pay the premium of call you are buying is (partially) from the premium you are collecting from selling put. The cost is very low, which provides leverage of trading.
But it might not be a good idea in the past year, since you could simply secure some kind of bond yield around 8%.
So if (big assumption) annual log-returns are iid normal with known mean m and variance v, then the long-term return after T years is also normal (call it X) with mean mT and variance vT. The capped ETF long-term return is sum of truncated iid normals with support (a,b) which not sure what it is (call it Y) but can be parametrized in terms of m,v,T,a and b. I wonder how E[X-Y] changes depending on the truncation bounds? For instance if m=5% and the bounds 0% and 10% are symmetrical around the expected annual return, then both index ETF and buffer ETF should return the same long-term. In practice, m~7-10% so the bounds are asymmetrical, which erodes the long-term return for the buffer ETF.
I think you're asking a good question -- whether the buffered ETF is a good deal depends on the bounds, so it would be interesting to see for what values it's a good deal or not. But you don't need to assume that the distribution of returns is normal or iid. You can use whatever model of returns you want, and compute the results by simulation. That way you don't get just E[X-Y], you can get the whole distribution.
Good article. The idea of rip off ETF coming from option trading, buy call (K=current price) and sell put (K=current price*1.1) which will secure profit from 0 to 10%. The good thing about this trade is that it is partially self financing, meaning you pay the premium of call you are buying is (partially) from the premium you are collecting from selling put. The cost is very low, which provides leverage of trading.
But it might not be a good idea in the past year, since you could simply secure some kind of bond yield around 8%.
Anotha, banga.