This article first appeared in the March 2013 newsletter. Subscribe to get articles like this in your inbox.

As mentioned in the last newsletter and by like-minded analysts, current valuations point to an expensive US stock market relative to market history. To value the market as a whole, our preferred metric is the P/E ratio calculated using 10 years or more of earnings.

One of the most common arguments against this approach is that it ignores interest rates, especially today when they are at historical lows. Surely, we can’t ignore interest rates when we evaluate the stock market, so let’s take a look at how we can take them into account.

Two ways that interest rates could potentially affect the stock market are:

1. Interest rates do not change the inverse relationship between current valuations and future returns. When interest rates are low, investors are willing to accept lower future returns and therefore pay a higher price today, making the market look expensive. Low interest rates just provide a rationale for high valuations.
2. Interest rates do change the relationship between valuations and returns, so they need to be included when predicting returns.
   To test point #1, we can calculate the historical correlation between valuation ratios and interest rates. The following table shows the mild inverse relationship between the two, especially if 10-20 years of earnings are used. Low interest rates do tend to occur with high valuation ratios, but it’s not a strong relationship.

Although somewhat interesting, this alone is not very useful for investors. What we want to know is the returns we can expect based on current valuations and interest rates, which brings us to point #2.

Let’s see if including interest rates helps in predicting future market returns. The following shows returns predicted using P/E alone (blue) and using P/E with interest rates (orange). The actual inflation-adjusted returns earned are in yellow.

Source: Robert Shiller data with Mariposa calculations, as of 3/27/2013.

The blue and orange lines are nearly the same, indicating that return estimates don’t depend on whether you include or ignore interest rates. The more important factor remains current valuations.

This article first appeared in the December 2012 newsletter. Subscribe to get articles like this in your inbox.

Valuation ratios such as P/E are not very useful in making one year predictions, although they do a fair job of predicting longer-term returns of at least 5-10 years.

Considering this, we will not perform a market forecast for 2013 or any other single year, unless we can support it with a high standard of evidence. Instead, let's use those valuation ratios to evaluate the markets today and estimate how that affects returns over the next 10 years. It’s not as exciting, but it’s far more responsible and reliable.

The following table shows valuation ratios for global stocks and US real estate as of 12/14/2012, with all P/E ratios using 10 years of earnings. Estimates for future returns assume that valuations revert to fair levels 10 years from now and that asset classes otherwise perform in line with their historical averages.

The two valuation ratios for US equities (Q and P/E) are calculated using separate sets of data yet produce similar results, indicating that our methods are consistent.

US stocks and especially US real estate are overvalued by a significant amount, so we can expect lower than historical average returns for both assets classes over the next 10 years. Valuations for stocks outside the US look far more attractive, making higher than average returns more likely.

Since predicting returns even in the 10 year range exhibits a high margin of error, tactical bets should be measured and sized appropriately based on investment goals and risk preferences.

This article first appeared in the September 2012 newsletter. Subscribe to get articles like this in your inbox.

When thinking about restricted stock units (RSUs) in a portfolio, most consider the tax and risk consequences of holding and selling restricted stock: should you elect to pay income tax when shares are granted rather than when they vest? How fast should you sell once they do vest? Should you use options to manage the risk of depreciation?

These are all important decisions, but they are already discussed at length elsewhere. Instead, let’s discuss how RSUs can affect your strategy for the rest of your portfolio.

Let’s say after careful consideration, you’ve decided on a 70% stock, 30% bond portfolio, but half of your savings is already in RSUs that you cannot sell. If you ignore your RSUs while managing your brokerage accounts, you essentially have two concurrent, but separate investment strategies. The risk profile of your overall portfolio is unfortunately largely dictated by your mix of RSUs vs brokerage accounts.

To manage both under one comprehensive strategy, you can adjust RSU balances by its risk attributes, tax liability, and vesting schedule to integrate it with an asset allocation framework.

Risk Attributes

First, determine which asset class is most similar to your restricted stock. For most, the appropriate asset class would be US stocks. If you use more narrow asset classes, it could be something like US value stocks or US small cap stocks. The asset class you select here would be most affected by your RSU balance.

Next, estimate how risky your stock is relative to that asset class. Smaller, growth companies tend to be more risky than larger, dividend-paying companies. An easy way to estimate this for public companies is by looking up the beta of your stock on Yahoo or Google Finance.

Tax Liability

Unless you specifically elected otherwise, income tax is due when shares vest. In this scenario, you can discount your unvested shares by your tax rate to estimate the after-tax value. For example, assuming a 40% total tax rate, your unvested (and untaxed) shares would count at 60%. If you did make the tax election, you should only discount for capital gains taxes.

Vesting Schedule

Similar to discounting future cashflows, you can discount shares by their vesting date. This accounts for the possibility that you are not with your current employer when shares are scheduled to vest. If we use a 25% discount rate for an example (you switch jobs every 4 years or so), shares due to vest in one year would count at 75%. Shares due to vest 2 years from now would count at roughly 56%, and so on. This calculation allows your shares to accrue smoothly to 100% as the vesting date approaches.

An Example

Let’s say you have $1mm in restricted stock (a small public company in the US) and you have $1mm in a brokerage account. And your target allocation for your overall portfolio is 35% US stocks, 35% non-US stocks, and 30% bonds. Let’s go through each adjustment:

• Asset Class: US stocks
• Risk Level (beta): 1.1x
• Tax Rate: 40%
• Vesting Schedule: all shares vest in 2 years
• Vesting Discount: 25% per year

The adjusted balance of your RSUs would be $1mm * 1.1 * (1-40%) * (1-25%)^2 or about $371k. For asset allocation purposes, your portfolio now looks like one worth $1.37mm ($1mm + $371k). Next, let’s look at two approaches to including RSUs in an asset allocation strategy.

Approach 1: Perfect Substitution

Based on your target asset allocation, your US stock allocation should be $480k (35% of $1.37mm). Since you already hold $371k in US stocks in the form of restricted stock, you only need to hold the remaining amount ($109k) in your brokerage account. You can use the rest of your brokerage account to hit your targets for other asset classes.

The downside of this approach is that most of your US stock allocation is now taken up by a single stock. Since that stock is not a perfect proxy for the whole asset class, we would be giving up some diversification benefits by doing this. And what if your RSU balance were much higher? Should you reduce US stock holdings in your brokerage account to zero? Probably not.

Approach 2: Balanced Substitution

A second approach attempts to balance the need to account for restricted stock for asset allocation purposes and the diversification benefits of holding a US stock fund even with a concentrated holding in a single US stock.

Unlike the first approach, as your RSU balance increases, you no longer reduce your US stock allocation 1:1. Instead, you start to reduce similar asset classes as well to preserve some balance in your portfolio. For an employee of a US company, you would reduce allocations in your brokerage accounts in the following order:

1. US stocks
2. Non-US stocks
3. Other risky assets, such as real estate and commodities
4. And eventually, even bonds, since they are more risky than cash

This approach satisfies multiple requirements in managing RSUs:

• Your overall portfolio is guided by your target allocation, since this approach does not ignore your RSUs
• Diversification is largely preserved, since an asset class would not be solely represented by a concentrated holding in a single stock
• The effect of RSUs is adjusted for risk, tax and vesting factors in a smooth, quantifiable way

Still have questions on how to incorporate RSUs into your asset allocation strategy? Feel free to contact me to discuss the details.

This article first appeared in the June 2012 newsletter. Subscribe to get articles like this in your inbox.

Over the years, my clients and I have developed strategies to use 529 plans effectively when building a college fund. Here are three to get you started.

Vanguard’s 529 plan is a good default choice if your state does not offer tax benefits for using an in-state plan. Most states (including CA, WA, HI, TX, and NJ) do not offer a tax benefit for in-state plans, so there is no opportunity cost with using an out-of-state plan like Vanguard’s. Of the many plans across the country, Vanguard’s plan is well-regarded for its low cost and variety of funds.

For New York residents, NY’s 529 plan is a great choice since it offers low-cost Vanguard funds while maintaining NY state tax benefits.

Open a separate account for each parent. Both plans mentioned above (Vanguard’s and NY’s) limit the number of investments at 5 and require at least a 5% investment in each option. By opening 2 accounts for one child, you essentially get to hold 10 investments and can allocate as little as 2.5%.

Also, 529 plans limit you to one trade per calendar year. With separate accounts, you can use this annual trade at different points in the year, giving you more flexibility.

Invest on a regular basis. Not just because this makes saving for college more manageable, but because both recommended plans let you make an investment selection for each deposit. Changing this selection with every deposit does not count as your annual trade. So if you make a monthly deposit, you have a chance every month to rebalance.

This article first appeared in the March 2012 newsletter. Subscribe to get articles like this in your inbox.

With Yelp, LinkedIn, Zynga, and Groupon completing IPOs in the past year and Facebook on the way, interest in IPOs seems to be surging. As with any investment strategy, my first reaction is to look for data. And not just data on a handful of recent IPOs for scenario analysis, but data on all IPOs over decades of market history.

The effect most investors are familiar with is that shares tend to appreciate significantly on the first day of trading. This is fully supported by evidence.

From 1980 to 2011, the average gain on the first day of trading is about 18%. Even if you remove the bubble years of 1999-2000, IPOs still average a return of 12% on the first day. This return obviously varies by company, but buying IPOs at the offering price seems to be a very profitable strategy.

Typical Investors

Unfortunately, buying an IPO at the offering price is really difficult if not impossible, since most shares are sold to large clients of the underwriters. For most investors, your first chance of buying an IPO is on the first day of trading–after the shares have already appreciated.

So how about buying shares after the IPO?

From 1980 to 2010, the 3-year performance of IPOs bought on the first day has lagged the broad market by an average of 20%. Although this under-performance disappears for companies with significant revenues (more than $50mm in 2005 dollars), buying IPOs on the first day does not look like a profitable strategy.

Employees

As an employee of a company completing an IPO, you are in the opposite situation of already holding shares and looking to reduce your position for risk management purposes.

Considering the performance record above, selling all unrestricted shares as soon as possible seems to be a decent default strategy, especially for employers with low revenues. However, making large, abrupt portfolio changes incurs significant risk. It can be particularly painful to see shares appreciate rapidly after selling a large block.

A common solution is to sell incrementally over a period of time to reduce this risk–usually over a few months to a few years. The length of time can be adjusted based on personal factors: your tax situation, the amount in employer stock relative to your overall portfolio, your desire for full diversification, your employer's growth prospects, your influence over company performance, etc.

From the December 2011 newsletter:

As mentioned last year, valuation ratios such as P/E are not very useful in making one year market predictions. This is disappointing, since they do a fair job of predicting longer-term returns of at least 5-10 years. As a result, you will not see a market forecast from us for 2012 or any other single year, unless we can support it with a high standard of evidence.

The next best thing is to evaluate the markets today using those valuation ratios and consider how that affects future returns in the 5-10 year time frame. It’s not as exciting, but it’s far more responsible and reliable based on historical market data.

The following table shows valuation ratios for US stocks and US real estate as of 12/23/2011, with P/E ratios using 10 years of earnings. Estimates for future returns assume that valuations revert to fair levels 10 years from now, and that asset classes otherwise perform in line with their historical averages.

The two valuation ratios for stocks (Q and P/E) are calculated using separate sets of data, yet produce essentially the same results, indicating that our methods are at least consistent.

Both stocks and real estate are currently overvalued by a significant amount, so we can expect lower than average returns for both assets classes over the next 10 years. However, tactical bets should be measured, since 10 year returns are still expected to be positive and the margin of error in these predictions is meaningful.

From the September 2011 newsletter:

In prior newsletters, I made the case for using the P/E ratio to value the stock market and to predict future returns. Now let’s apply the same analysis to real estate, or more specifically REITs.

One crucial feature of REITs is that they are required to distribute almost all their income to maintain their tax status. This means earnings (the E in P/E) can be closely approximated by dividends. And dividend data is far easier to find for individual investors. Even better, the REIT industry provides historical data on their website for budding data scientists.

With a little data manipulation, you can create the chart below. “P/E” in this case uses 10 years of dividends, not earnings. The relationship we saw for stocks seems to hold again for real estate: P/E ratios using a longer horizon have a close, but not perfect, relationship with subsequent returns–in this case 2 year returns. High P/E ratios (blue line) tend to line up with below average returns (red line).

Source: REIT.com data and Mariposa calculations as of 9/28/2011.

Like stocks, this relationship does not work as well at shorter time periods (1 year or less). And below average returns for real estate can still be positive, weakening the case for using this analysis for short-term trading. As a result, my primary use is to make measured tactical adjustments vs strategic asset allocation targets, knowing that it may take years for the market to correct.

And how does the US REIT market look today? At Monday’s close (VNQ at $56.06), REITs look to be around 40% overvalued. It’s not alarming, but you should expect returns that are lower than the historical average.

From the June 2011 newsletter:

If you have ever switched jobs, you probably have more than one old 401k (or 403b) account with past employers, with the occasional unexpected statement as your only reminder that they even exist. You suspect just leaving it there is not what you should be doing, but what is the best strategy with old 401k accounts?

When leaving a job and starting a new one, you generally have four options with your old 401k (exceptions always apply):

1. Leave it there.
2. Transfer it to your new 401k.
3. Roll it over to a Rollover IRA.
4. Roll/convert it to a Roth IRA.

What do I do? I’ve generally rolled my old 401k accounts to a Rollover IRA as soon as I leave. I do this mainly for the freedom to buy the investments that I want, which are usually lower-cost and a better fit for my retirement portfolio. Also, I keep the option to convert to a Roth in the future. However, there are good reasons to choose some of the other options.

Especially if your 401k is your sole investment account, transferring your old one to your new job keeps things very simple. You only have to worry about managing one account. Since you are limited to the investment choices in your new 401k, this is a good option for simplicity as long as those new investment choices are decent.

An advanced option is to roll your old 401k to a Roth IRA (doing a Roth conversion in the process). Since this is just a combination of rolling it over to a Rollover IRA and then converting it to a Roth, you should only consider this if you’ve decided to do both steps anyway. Remember that doing this triggers income taxes, just like a regular Roth conversion would.

And if you are still undecided, leaving it there is a perfectly fine strategy as you figure out what to do next. And it’s also a good temporary option if you’re just waiting for your new job to start, or waiting for the next calendar year to do a Roth conversion.

Another factor you may consider is that 401k’s and IRAs can have differing asset protection features. If this is a critical feature for you, make sure to consult an attorney.

One option I intentionally left out is to withdraw the funds, incurring income taxes, and penalties for those less than retirement age. Although this could be an emergency source of cash, it is not a recommended option. Really, do one of the other options.

From the March 2011 newsletter:

When I list “full diversification” as a core component of Mariposa’s investment strategy, what does that really mean? What’s the point of diversification?

Before answering those questions, let’s first discuss one interesting, but often overlooked, property of investments: the difference between the realized, annualized return and the average return. For example, if you experience returns of +10% in one year and then -10% in the next, your realized return is not 0%, but slightly negative (-1% if you’re calculating at home). I like to think about this relationship as a simple formula:

Annualized Return = Average Return - Risk Penalty

In the example above, the average return is zero, but the annualized return (the one that matters) is slightly negative because there is some risk. If the risk were higher (let’s say returns of +50%, then -50%), then the risk penalty would be higher, reducing the annualized return further.

Knowing this, one way to increase the type of returns that you care about (annualized) is to reduce risk without reducing average returns. The advantage of diversification is that it does just that.

Let’s look at a very simply diversified portfolio composed of 60% US stocks and 40% US bonds. This example works because bonds and stocks generally don’t move up or down together. In many cases, they move in opposite directions, absorbing some of the shock of market movements.

The total portfolio earns an annualized return that is 0.54% higher than what you would expect if you simply averaged the components. This “bonus” in annualized return is purely due to the reduction in risk, not any increases in average returns. Diversification in this case lets you capture more of the average return that is rightfully yours.

If diversification works with a very simple 60-40 portfolio, wouldn’t you want to try portfolios that are more fully diversified? You can easily add investments such as real estate, foreign stocks, and commodities to diversify and reduce risk even further. And these are the results if you do that, while still keeping the overall percentages as 60% risky assets (stocks, real estate, and commodities) and 40% bonds:

In this case, the diversification benefit is 1.40% per year. The addition of investments that move differently from US stocks and bonds provides a significant return boost, even if the average return doesn’t change.

So ignore active mutual funds, Morningstar ratings, or even earnings forecasts, just select an asset allocation that has the appropriate level of risk and diversification for you and rebalance periodically. Then spend the extra time and returns on activities you actually enjoy.

Here are the five most popular articles with our readers in 2010, our first full year of blogging:

1. Are Morningstar Ratings Useful?
   A brief evaluation of Morningstar ratings as a predictor of future mutual fund performance. This article was also published on Get Rich Slowly.
2. Is Real Estate a Good Investment? (Part 1)
   The first of a two-part series covering real estate as an investment. This one evaluates buying rental property as an investment. Part 2 covers buying a home vs renting.
3. Glide Path: A Target-Date Fund’s Secret Sauce
   Although marketed as a all-in-one solution, not all target-date funds are alike. One difference is how each fund changes its asset allocation over time (its glide path).
4. Online Investment Advisors: Marketriders, Betterment, and Plantly
   New online investment companies are bringing customization and automation to personal investing. See how they match up.
5. Top 3 Problems with 529 Age-Based Options
   Although a decent all-in-one option for 529 plans, there are a few design flaws to be aware of with age-based options.