Infinite Horizon wrote: ↑Mon Jun 15, 2020 9:06 pm
I talked this out with an economics person, and they said that if margin constraints bind hard, I should discount with my risk-adjusted leveraged returns (so ~7-9% rather than ~0%). If I go above Kelly leverage I go to near 0 with high probability, and I can't pull out the illiquid assets to reboot early. However, I should be very enthusiastic if I do find the opportunity to borrow against the illiquid assets/capitalized wages, or take an unsecured loan at a reasonable interest rate.

I don't think that person understands this concept very well. Imagine this scenario:

You have 100k in savings. The present value of future savings, discounted with RFR is 900k. Discounted at 8%, say it's 300k (a factor of 3 like you mentioned). You want a 50/50 stock/bond allocation. So if you follow the Merton model, you'd want to invest 500k in stocks today, or 5:1 leverage. You can't do that due to credit constraints. So you'd follow the Ayres advice and just do 2:1 leverage. So then you say "OK, credit constraints mean 2:1 leverage is the best I can do. And based on this friend of mine, if I'm leveraged constrained, then I need to discount at stock market rate". So then your portfolio is really 100k+400k and you'd need to invest 250k in the market today.

**Which you still can't do any ways since the most you can do is 2:1 leverage.**
So guess what, both of these result in leveraging to the max any ways

**so discounting at a higher rate in no way addressed your leverage constraints**.

A sure payment next year is worth whatever treasury zero coupon bond you need to buy today to pay that in a year, period. Using a higher discount rate reflects the uncertainty of that payment. To the extent your job is safe and you know you will save in the future, one way or another, these payments should be, indisputably, discounted at the RFR. If your job is high risk, then you discount somewhat higher. If it's as risky as the stock market itself (some years you get paid 20% more, some years you have to pay 50% of your wage, etc), then it makes sense to discount with the stock market rate.

The way to address leverage constraints, as I said before, is to put a cap on the leverage you use.

As to your friend, remind him that the Kelly Criterion of the market historically (about 1.5x leverage) only applies in the absence of fund inflows and outflows. If you are planning to contribute to your account, the Kelly Criterion increases significantly, because it is akin to the historical Kelly except every scenario comes with additional payments. So you can leverage more, optimally, if you contribute to an account. Same the other way; if you're in retirement and withdrawing, you can't come anywhere near close to the historical Kelly Criterion of 1.5x leverage without risking ruin.

Remember that every retirement cohort since 1887 in the US, Japan and Britain, would've retired with more money following this strategy? And in the simulations, the cohorts leveraged 2:1, "couldn't pull out illiquid assets", and went past the Kelly Criterion.

"... so high a present discounted value of wealth, it is only prudent for him to put more into common stocks compared to his present tangible wealth, borrowing if necessary" - Paul Samuelson