@jkandell Good catch! Leaving the box unchecked means that the returns are 100% independent of each other (we use a unique random number for each asset class in generating the randomized returns). I'll clarify this in the next update of the user manual.
Stuart
Happy to help. You mean you're not using Cholesky decomposition to generate random numbers with known correlations? (I'm being sarcastic.)
For the benefit of users deciding whether to leave that checkmark blank or to check it, given that the historic correlation between TSM and BND has been +0.20 (+.30 since 1970), and TSM and T-bills +0.12, the "100% independent" (ie average of 0% correlation) is much closer to historical reality than the checkbox option (=100% correlated). So I personally would never check it now that you've explained what it does.
Also we must remember that any potential diversification benefit is easy washed away by failure to rebalance when the time comes, or simply having such small allocations that it doesn't move things, and/or trading costs.
In short: if you don't have substantial investments in bonds/t-bills, Pralana's results will roughly model true expected returns, at least in the 25-75th bands. In the case (a) your portfolio is mostly Bonds, especially T-bills, with 15-30% stock, and (b) you rebalance regularly (that is you have the resolve to actually shift funds after a crash to a risky investment that has done terribly), the default checkbox option (unchecked) will slightly underestimate your expected returns, at least in the 25-75th percentile bands.
