Allocating my assets between mutual funds in my IRA account

February 17, 2009 8:06 AM

I have four mutual funds in my IRA account - a Total Market index fund, a Small-cap index fund, an International stock index fund, and a Bond index fund, and I'm trying to figure out how to allocate my assets between them. I've read a little bit about "Modern Portfolio Theory" and the "Efficient Frontier", but I'm struggling to understand some of the math. So, is there a simple way I can test whether a particular allocation is on the efficient frontier? Or, see all the possible allocations on the efficient frontier and choose between them? How do I figure this out? Ideally, I'd love to see a simple enough formula that, given the mean and standard deviation (and maybe correlation matrix), of my four funds, would tell me what allocation is on the efficient frontier. Or, some type of online tool, or an Excel spreadsheet or something.

As an aside, I'm not interested in using a financial advisor (I don't have nearly enough assets to justify that), nor am I interested in general rules of thumb ("If you're N years old, you should have an allocation of x% stocks, y% bonds, etc.")

Nothing wrong with that - just seems like this is something I should be able to understand myself.

A lot of what I've found on Google is high-level explanation (e.g., what MPT and EF are), but not the tools to actually allocate between N funds. Or, very sophisticated math that I don't understand.

Of course, maybe I'm wrong - maybe the math is intractably hard. But, maybe not!

Many thanks in advance for any advice or insight!

posted by stuehler to Work & Money (10 answers total)

Do you have Excel? Solver.com's Investment Examples is exactly what you need, Markowitz portfolio optimization.

I don't know if I'd throw my assets into a blackbox I didn't understand (and of course the aforementioned examples use gaussian randomness yadda yadda), but that should do ya. You can do pretty much all of modern finance with Excel and VBA, which is so simple to learn.
posted by geoff. at 8:36 AM on February 17, 2009

Response by poster: geoff,

Many thanks for the quick reply - this looks like exactly what I need!

However, I can't figure out that Solver.com site. I have the Solver add-in installed. But I don't see how to download the workbook you linked to.

I'm sure I'm missing something obvious.

Any pointers?

Cheers,
Matt
posted by stuehler at 9:15 AM on February 17, 2009

I'm going to caveat all this by saying I think this is a tremedous waste of time. By definition the variance of returns will be backward looking. That said its a pretty simple if long formula to figure out the standard deviation of returns, and you should already have a sense for what your expected long-term returns are for each asset class then you can just figure out using goal seek in .xls what the optimal formula is. I'd also argue that expected returns aren't static - i.e you could surmise equity returns going forward should be higher then history, fixed income returns lower - as they mean revert to long-term averages.

To figure out the standard deviation of a portfolio follow the math on this page. The math itself is easy - you will need to source the correlations from somewhere. I'm sure a little googling will get you the answer. I'm pretty sure given how common the four fund types you have chosen are you should be able to find what you need.
posted by JPD at 9:23 AM on February 17, 2009

Response by poster: JPD,

I appreciate your caveat - I don't have any expectation that using a method like this to choose an allocation is any sort of guarantee.

But, I've put a lot of time and effort into choosing the funds (by focusing primarily on expense ratios). Now, since you can't NOT choose an allocation, I've got to do something. So, my options are to choose something completely random, distribute the assets equally, use some general rule of thumb gleaned from a site like the Motley Fool, or, I guess, use something like a model based on MPT/EF.

I figured that the last method, assuming it's easy enough, is probably my best bet. Are you suggesting that there's a better option?

In any case, I definetly appreciate your other pointers. Many thanks!
posted by stuehler at 9:36 AM on February 17, 2009

Ask this at diehards.org, a site for diy investors. It's not sleazy, as many other financial sites can be. There's a lot of knowledgeable people there, and many are spreadsheet pros. There's a lot of discussion about asset allocation, the efficient frontier, and so on.
posted by hardcore taters at 9:39 AM on February 17, 2009

Well I think the whole notion of return variation = risk is silly so that right there means I think the portfolio theory/ mean-variance approach to portfolio creation is pointless - but that is a discussion for another time.

In the context of the question you asked - coming up with a systematic approach to portfolio weighting this is probably your best option. A professional would probably do the same flavor of math, just on a different scale.
posted by JPD at 10:09 AM on February 17, 2009

The whole area of risk and return in portfolio management can be pretty complex, as you've probably noticed. My feeling is that I would caution against trying to construct your portfolio in this manner unless you fully understand the concepts behind it. Which means being able to fully understand the math and theories involved on a conceptual level, which it sounds as though you may be struggling with. If you don't want to consult with a financial advisor, there is lots of good information out there on how to build your portfolio based on things like your risk tolerance, time horizon, goals etc. I'm a big fan of the articles on Investopedia and this one may be a good place to start at, if you're so inclined.
posted by triggerfinger at 11:02 AM on February 17, 2009

I agree with the warnings by JPD. MVO is a very dangerous tool. You speak about trying to find an allocation that is on the "efficient frontier" as if that is something that is fixed in time. It is only a result that is computed after the fact and it is extremely time dependent. It will only tell you what worked in the past. It should not be used to pick an allocation but as a learning tool that can give you a feel for how risk and return vary as you push the parameters around.

This is the MVO tool used by William Bernstein, author of The Intelligent Asset Allocator and The Four Pillars of Investing.

You might also look at these two links where a lot of the work has already been done.

Link1
Link2

Remember, the greatest enemy of a good plan is the dream of a perfect plan. The selection of your equity to bond ratio will determine 90% of you investment return. The rest is just fine tuning around the edges with no guarantee of success and a real risk of harm.

You certainly have the right idea -- four low cost index funds should be all that you need.
posted by JackFlash at 3:33 PM on February 17, 2009 [1 favorite]

Definitely read up on the EF and asset allocation at diehards.org - lots of links and informative threads there, it's a group of people who (mainly) are inspired by the founder of the Vanguard index funds.

Also, Morningstar has an online tool called Instant Xray that lets you enter your funds and $$ amounts and tells you your overall exposure to different asset classes, sectors, etc. Can be useful for deciding the right amounts to put in each fund, once you've figured out your target allocations.

But JackFlash is right: don't over think things, the difference between the "best" strategy and a "pretty good" strategy isn't too huge in terms of overall returns.
posted by Fin Azvandi at 6:52 PM on February 17, 2009

This:
Remember, the greatest enemy of a good plan is the dream of a perfect plan. The selection of your equity to bond ratio will determine 90% of you investment return. The rest is just fine tuning around the edges with no guarantee of success and a real risk of harm.


I wish I had said this.
posted by JPD at 6:57 AM on February 18, 2009