How do I model total capital loss using Monte Carlo?
I want to model my stock/bond ratio changing over time. E.g. I start off with 70/30 and expect to end with 0/100. I model this by using Monte Carlo and creating a portfolio for each decade or so with the set stock/bond ratio for that decade.
However a scenario I want to model is essentially the equivalent of upside investing. Which is, what happens if all my stock money (including the principal) is lost at a specific point in time? I would model this by specifying an investment in one of the portfolios that has effectively a -1 return.
But I've never been able to enter any values for a custom asset without E$Planner losing it's mind. Can someone give me asset settings that will work with a -1 return? I don't care what the variance of correlation is since I ignore everything but the median anyway.
In other words I only use Monte Carlo because it lets me have portfolios, not because I think Monte Carlo simulations are useful for modeling stock market behavior.
Subscribe to RSS feed
Comments
dan royer
Sun, 11/30/2014 - 10:17
Permalink
Because I have the same view
Because I have the same view as you with regard to Monte Carlo, I have not worked much with user defined assets. I know that our programmer is working right now on the UDA inputs and I will show him your question.
I note that using economics mode, you have both Husband and Wife assets (that's two pools of money) and you have the option in the Assumptions area to change the future rate of return on each of those pools once. So that's four rates of return that can be modeled. That's probably not enough to address your need here. But of course you can clone the profile and run it again a second or third, or fourth time with different nominal rates of return. That's probably what I'd do for now: run the profile multiple times using various sets of nominal returns and ignore the MC.
esplanner@goland.org
Tue, 12/02/2014 - 11:51
Permalink
The problem with the
The problem with the assumptions approach is that the size of the various pools of money (e.g. regular assets, husband retirement assets and wife retirement assets) are very different. So I'd have to play some fairly unpleasant games with calculating effective return rates across time across all three asset areas to try and simulate something useful. I'm not sure it's even possible to get an accurate result in our case. If we just had retirement assets, perhaps. But we have non-trivial regular assets as well.
The approach I'm using now is that I set up 5 MC portfolios to simulate different returns for different combinations of stocks and bonds that we would switch to as we increase our bond holdings. Since I can't figure out how to enter user defined assets without causing a mess I instead use 'synthetic' assets. For example, I calculated that if I wanted to simulate a 0% return (which isn't exactly what I want since it doesn't include loss of capital) then I should assign 16.67% of the money with 0% return to Cash and the rest to Short Term Government Bonds. The median return is then 0%.