Strategies

Does an asset class allocation of less than 5% make sense?

In allocating the asset classes in a portfolio, many “experts” recommend that to keep things simple, it does not make sense to sub-divide the portfolio into anything less than 5%. The usual cited reason is that it complicates the portfolio with very little increase in benefit.

For example, if an asset class has an allocation of 2.5%, even if it were to gain 20%, its overall impact on the portfolio is going to be only 0.5%.

Recently, this question came up on the Bogleheads’ forum Why allocate less than 5% to anything? Roy crunched some numbers for illustration:

Since 1972: 100% S&P 500:
CAGR 9.26%
Standard Dev 18.59%

Since 1972: 97.5% S&P 500 and 2.5% CCFs:
CAGR 9.35%
Standard Dev 18.17%

Since 1972: 95% S&P 500 and 2.5% CCFs and 2.5% REITs:
CAGR 9.43%
Standard Dev 17.97%

There appears to have supporters on each sides. Yet, the resident experts both seemed to be OK with the idea of less than 5% allocations:

Larry Swedroe has about 3% in CCFs in his personal portfolio and Rick Ferri suggests only 2% or so in microcap.

For me personally, my current minimum allocation is 4%. Larry sums up this issue with the following statement:

It is not an issue of doing much, if the cost is effectively zero if it adds anything it should be done.

Squeezing out more tax efficiency from my portfolio

My portfolio is currently allocated as follows:

  • US LargeCap 11%
  • US SmallCap 8%
  • US REITs 4%
  • US Timber REITs 4%
  • US MLP (PIK dividends only) 4%
  • INT EAFE 20%
  • INT Emerging Market 11%
  • PreciousMetal Equity 5%
  • Fixed Income 33%

The ranking of the asset classes, in terms of tax efficiency (based on historical dividends), are as follows (most tax efficient at the top):

  • US MLPs EEQ & KMR (with PIK paid-in-kind dividends only) : 0% taxed.
  • US Timber REITs PCH, PCL & RYN: 0% taxed due to my enormous capital losses accumulated (see related post “Tax-Free Dividends from Timber REIT Investing” HERE).
  • US SmallCap (IJS) : distribution yield ~2.0% will be taxed.
  • PreciousMetal Equity (GDX) : distribution yield ~2.0% will be taxed.
  • US LargeCap (IWB) : distribution yield ~2.8% will be taxed.
  • INT EAFE (VEA) : distribution yield ~4.0% will be taxed.
  • INT Emerging Markets (VWO) : distribution yield ~5.0% will be taxed.
  • Least efficient : US REITs and Fixed Income.

The commonly sprouted ‘wisdom’ of putting International equities and emerging market equity in the taxable account so as to get the foreign tax credit is coming back to bite me now (and probably in future as well) with huge 4% (VEA) and 5% (EEM) distribution yields. This is ironic considering that VEA is simply just another class of Vanguard’s Tax-Managed International stock fund.

On the other hand, US SmallCap Value has a sub-3.0% yield but ‘experts’ usually would recommend putting this asset class into tax-deferred accounts.

Anyway, based on this analysis, I have been re-directing new money in the taxable account to the most tax-efficient assets listed above.

Running out of tax-deferred space

I am running out of tax-deferred space for my investment portfolio. But before going into the problem, here’s some background information.

Background

For tax reasons, it is desirable to place each asset class into the appropriate “investment location” (by “investment location”, I mean whether it is in the traditional IRA, Roth IRA or taxable account). To understand why this is so, we need to first understand the concept of an asset class’s tax-efficiency.

An asset class which throws off a lot of taxable income is considered to have low tax-efficiency. This is because if it is placed in a taxable account, a lot of the return, usually in the form of taxable income, will be taxed (at federal income tax rates of up to 35%  plus state taxes, if applicable) and taken away by the tax man. Asset classes like taxable bonds and REITs comes to mind.

Conversely, an asset class that mostly appreciates in value (capital gain) and gives very little income or capital gains distribution is considered tax-efficient. It is tax-efficient because you do not have to pay any (or very little) taxes on it for as long as you continue to hold the investment. The total stock market index fund would be such an asset class.

If the asset class placement is incorrect, an investor may have to pay significant ongoing taxes on the investment, reducing the total return. Ideally, tax inefficient asset classes should have priority placement in tax-deferred accounts; and tax-efficient asset classes can be kept in the taxable account.

To read further on this, the Bogleheads Wiki has a topic on this issue. The following is a chart taken from the Wiki:

Problem

In my portfolio, I have three main tax-inefficient equity asset classes / vehicles. These are

  1. US REITs : Vanguard REIT index mutual fund (VGSIX) and ETF (VNQ).
  2. Precious Metals and Mining : Vanguard PME fund (VGPMX).
  3. International SmallCaps : Vanguard International Explorer fund (VINEX).

(Note: the rest of my portfolio is shown HERE).

I am currently holding these funds in tax-deferred accounts with Vanguard and Wells Fargo. However, as the portfolio gets larger (especially on the taxable account side), the tax-deferred space for holding these investments is shrinking (in relative terms).

Resolution

I have basically two options: (1) to reduce my allocation to the tax-inefficient asset classes / vehicles or (2) to find tax-efficient vehicles and shift a port of the asset class to the taxable account.

I decided that option (2) is appropriate for me. I set out to look for ETFs which can be asset class substitutes and are also likely to be tax-efficient enough to be held in the taxable account.

For US REITs, clearly there is no good substitute because REITs by law have to distribute their income and so are, by nature, tax-inefficient. (Note: some have suggested that Third Avenue’s TAREX is a mildly tax-efficient substitute).

For the Vanguard PME fund VGPMX, XME (SPDR S&P Metals and Mining ETF, expense ratio 0.35%) is a close substitute. It’s expense ratio is low and has good trading volume. However, my recent experience with it showed that it is extremely volatile, moving perhaps two or three times as much as that of VGPMX on a daily basis. Below is a chart of their price performances for the past six months.

I wanted a fund that behaves similarly to VGPMX, but not two times as volatile! I decided to look further.

For international smallcaps, there are a variety of ETFs that can probably serve as good substitutes. The more promising ones include DLS (ER 0.58%), GWX (ER 0.59%) and SCZ (ER 0.40%). (For a complete list and description, see the Bogleheads Wiki.) Below is the price performance chart of the ETFs versus VINEX for the past six months.

I think I will skip DLS because it is weighted by dividend stocks and  theoretically less diversified than the other two because stocks will no or low dividend yields are excluded from its holdings. It would probably also be less tax-efficient because of its higher dividend yield.

Between GWX and SCZ, I am tempted to pick SCZ because of its lower expense ratio. However, its trading volume is also lower. GWX has a higher trading volume but its expense ratio is also higher. In the end, I think I would probably end up shifting a portion of VINEX to the both of them, and using them interchangeably for tax loss harvesting purposes.