Archives for July 2017

A reader writes in, asking:

“As I try to select among the funds in my 401k plan, is there a limit to the high end of expense ratio that you suggest not going beyond?”

The maximum expense ratio that I would pay for a fund depends on the context. Specifically, it depends on the availability of a less expensive suitable substitute.

For example, consider an investor with a $100,000 portfolio, all of which is in her 401(k). She has decided that her ideal asset allocation would be:

• 60% in US stocks,
• 20% in international stocks, and
• 20% in bonds.

And, in her 401(k), the lowest-cost investment options in each of those categories are:

• A “total stock market” index fund with an expense ratio of 0.1%,
• An actively managed international stock fund with an expense ratio of 0.7%, and
• A “total bond market” index fund with an expense ratio of 0.1%.

Our investor has a dilemma. She wants a 20% international stock allocation in her portfolio. But she would have to pay an extra 0.6% per year for that part of the portfolio in order to have such an allocation (rather than just allocating the entire stock part of the portfolio to the US stock index fund).

Frankly, if I were in that position, I wouldn’t pay the additional cost. I appreciate international diversification, but an avoidable cost of 0.6% per year is a high hurdle for the diversification to overcome in terms of added value.

But what if our investor’s lowest-cost investment options were as follows?

• An actively managed US stock fund, with an expense ratio of 0.65%,
• An actively managed international stock fund with an expense ratio of 0.7%, and
• A “total bond market” index fund with an expense ratio of 0.1%.

In this case, our hypothetical investor would now only be paying an additional 0.05% per year for her international allocation. The international fund’s cost hasn’t changed at all, but it has become quite a bit more attractive because the alternative (i.e., keeping the entire stock allocation in the US stock fund) has become more expensive.

Point being: The maximum expense ratio that it makes sense to pay depends on the cost and suitability of the nearest substitute. In our first scenario the 0.7% expense ratio was too high because there was a low-cost substitute (albeit an imperfect substitute). In the second scenario, the 0.7% expense ratio was not prohibitive, because there was no low-cost substitute.

What if There Are Multiple Accounts?

Let’s look at one more hypothetical scenario. Our investor still desires an allocation of 60% US stocks, 20% international stocks, and 20% bonds. And the options in her 401(k) are the same as in our first scenario above:

• A “total stock market” index fund with an expense ratio of 0.1%,
• An actively managed international stock fund with an expense ratio of 0.7%, and
• A “total bond market” index fund with an expense ratio of 0.1%.

In this case, however, her portfolio consists of $50,000 in her 401(k) and $50,000 in a Vanguard IRA. In this case, she has no reason whatsoever to pay the high cost of the international fund in her 401(k), because she could achieve her desired overall allocation by buying an international index fund/ETF in her IRA.

Again, the overall conclusion is that the maximum “acceptable” expense ratio for a fund varies based on the availability of less expensive suitable substitutes.

One question I receive periodically is why the Target Retirement and LifeStrategy funds at Vanguard hold “Investor” shares of the underlying funds, rather than the less expensive “Admiral” shares. Emily Farrell, Head of U.S. Business PR at Vanguard, was kind enough to provide the answer.

—–

Piper: Why don’t the LifeStrategy and Target Retirement funds use Admiral shares as the underlying holdings?

Or, if there’s a concern about that being a way to “cheat” to get access to Admiral shares without meeting the requirements, why not have an Admiral share class for the LifeStrategy and Target Retirement Funds, with a higher minimum investment (i.e., a high enough figure such that a person investing that amount via an identical DIY allocation would qualify for Admiral shares of all or most of the underlying funds)?

Farrell: Vanguard is unable to offer multiple share classes on funds of funds due to the structure in which we operate, and the agreement we have with the SEC regarding multiple share classes. Our funds of funds have no direct costs — the costs are derived only from the ERs of the underlying funds. According to that agreement, we must be able to offer a differentiated cost advantage between share classes. (Therefore, because there is no direct cost associated with a fund of fund — it’s not possible for us to offer such an advantage).

As you may recall, we launched a suite of Institutional TRFs in 2015. Note, however, those are not share classes — they are an entirely new set of mutual funds.

Piper: Given that Vanguard can only have one share class of a fund-of-funds, how does Vanguard decide which share classes to hold within such funds? The “normal” (non-institutional) versions of the funds-of-funds each hold Investor shares of the underlying funds. Would Vanguard ever consider holding Admiral share classes in these funds?

And, for example, the Vanguard Institutional Target Retirement 2050 Fund (VTRLX) holds Institutional shares of one underlying fund, Admiral shares of another fund, and Investor shares of two funds. What’s the line of thinking there?

Farrell: It is largely based on the cost to serve the underlying shareholder of the TRF or LifeStrategy Fund. These funds tend to attract small balance accounts and IRAs, and the cost to serve is disproportionately higher than funds with large and diverse shareholders. In terms of looking into adding holding Admiral in the Investor suite of TRFs, in short, the answer is yes — we consistently review our products, product structures, and product line-ups. However, given the size of the average TRF account and our cost allocation methodologies, we would not expect a change from Investor Shares. Specific to the underlying holdings of the Institutional suite of TRFs, again, it is cost to service the funds and our cost allocation methodologies that dictate the share class utilized.

—–

So, to paraphrase, Vanguard’s position here is, “in a Target Retirement or LifeStrategy fund, we hold whichever share class (or share classes) result in an expense ratio that most closely reflects the cost to service the fund’s shareholders.” In hindsight, that feels sort of obvious, given Vanguard’s general “at cost” pricing philosophy.