When it comes to expenses, CEFs are quite different than open-end mutual funds.
Perhaps the most common complaint I hear from investors who have decided not to invest in closed-end funds is that they are so expensive relative to open-end fund investment alternatives. It's a good point, but it misses the other part of the question: What do you get for that expense?
This week, we continue our series on CEF basics by looking at expenses. We found that, at least for their most recent fiscal years, CEFs as a group are on average more expensive than open-end mutual funds as a group.
Blame It on Leverage
Why are CEFs typically more expensive?
For any fund, closed-end or open-end, management fees tend to be the highest expense on the year-end income statement. These fees come from the management contract between the fund and its investment manager. One cause for CEFs' higher costs could be that CEFs charge higher management fees than their open-end competitors. Anecdotal evidence does not support this--we often find that a CEF charges the same management fee percentage as a sister fund with an open-end structure, with one major difference discussed below.
Instead, higher CEF expenses come from two factors. First, most CEFs typically assess the management fee against total--not net--assets. In such cases, the management fee is also charged against any structural leverage (lines of credit or preferred shares, typically) that the fund has. Second, if a fund incurs interest expense (most likely, from a line of credit), that interest expense by law must be included in the CEF's expense ratio. This makes sense, because the CEF's common shareholders bear the cost of that expense. What does not make sense is that if a fund has leverage through preferred shares, the distributions to preferred shareholders do not have to be included in the expense ratio, even though the CEF's common shareholders still bear the cost of this leverage.
Three Illustrative Examples
The table above shows three fictitious funds and how the use of leverage affects the funds' expense ratios. All three charge a 1% management fee on total--not net--assets, and all three have $200,000 in other costs. The language around management fees tends to differ; some funds charge a percentage of average daily assets, and others assess weekly or even monthly average assets. For the sake of simplicity, we are assuming the asset amount doesn't change over the period in question. Obviously, given portfolio performance, this wouldn't be the case in the real world.
The first fund has no leverage. It's rather straightforward. The total costs of $6.2 million come from the management fee ($6 million) and other costs ($0.2 million). In this case, for the year, the total expense ratio (total costs/average net assets) would be 1.03%.
The second fund is simply the first fund, but we've added $300 million in leverage via a line of credit. The management fee is $9 million because we not only charge the net assets, but also the assets from the line of credit. There's interest expense here, too. Assuming a 1.25% total interest rate on the line of credit, the fund would bear $3.75 million in interest costs. Total costs for this fund would be $12.95 million, which would all have to be reported in the fund's expense ratio of 2.16%.
The third fund is just like the second fund, except that instead of utilizing a line of credit for leverage this fund issued preferred shares. Note that, just as in the second fund, the management fee is $9 million as it is assessed against total assets. However, unlike the second fund, the cost of leverage is significantly higher ($15 million versus $3.75 million). It is important to understand that this significantly higher leverage cost is not included in the total expense ratio. So for purposes of the reported expense ratio, the fund would have only $9.2 million in total costs for a total expense ratio of 1.53%. On the surface, this looks much better than the second fund's expense ratio. However, a truer economic picture of the third fund's total costs would include the preferred distribution, boosting the total costs to $24.2 million for a truer expense ratio of 4.03%.
Given the discrepancy between leverage costs for funds 2 and 3, you may be wondering why a fund would issue preferred shares instead of tapping a line of credit. Well, in reality, I was pretty tough on preferred shares in this example. Most preferred shares, especially from municipal bond CEFs, offer variable rates of interest that are much lower than 5%. This example sprang to mind due to Gabelli Equity's
There are two key things to take away from these examples. First, a fund's reported expense may not be representative of the true economic costs that shareholders bear. Second, it's nearly impossible to compare funds on the basis of reported expenses. For comparison, for funds that incur interest expense, we produce an adjusted expense ratio figure that strips out the interest expense; this allows for better comparison between unleveraged CEFs, CEFs leveraged with lines of credit, and CEFs leveraged via preferred shares, although for funds charging management fees against leveraged assets the comparison is still not perfect.
An Analyst's View of Leverage-Induced Expenses
A better way, in our opinion, to look at expenses for leveraged CEFs is to first back out expenses that can be directly attributed to leverage and then to ask whether the use of leverage was profitable over the period being considered.
In this table, fund A uses leverage from a line of credit. Between the management fee incurred on leverage and the interest expense for the credit line, the total cost of leverage is 2.25%. Now, for total return on leverage, we simply use the fund's net asset value, or NAV, total return, because there's no way to back out the assets bought via leverage. With a 4.00% total return from investing, including the leveraged assets, this fund gained 1.75% on its leveraged assets over the year. So, yes, the leverage cost shareholders money but the net result was a gain.
Fund B uses preferred shares for leverage. In this example, the total cost of leverage is 6.00% and, as with fund A, the portfolio had a total return of 4.00%. In this case, Fund B lost money using leverage over the course of the year. In fact, it lost 2% on the leveraged assets.
It's important to bear in mind that one year isn't a very good yardstick to use for long-term investing. Odds are, as we'll see later, that over time fund B would overcome the high hurdle rate on its leverage. However, it is critical to understand that the cost of leverage is very important to investors, and almost universally to the fund managers I speak with. The lower the cost of leverage, the more likely it is that the leverage will deliver positive returns for shareholders.
Are CEFs Really More Expensive?
Comparing funds' expense ratios is a difficult undertaking. It can be imprecise, to say the least. Still, broad trends can be discerned, and it's a worthwhile endeavor to try to get our heads around the costs of owning a CEF versus the costs of owning a mutual fund.
For our study of funds' latest reported expenses, we looked at asset-weighted and equal-weighted results. Asset-weighted expense ratios give more sway to larger funds, in the belief that more investors are in those funds. Equal-weighted expense ratios give equal weight to all funds and are indicative of the range of choices available to investors. For CEFs, we looked at the total expense ratios and the adjusted expense ratios, where we stripped out interest expenses. We then grouped all of the funds according to their Morningstar categories and pitted averages against averages. All CEFs fall into 58 different categories.
Looking at asset-weighted category average comparisons (click here for results), it's quite clear that CEFs do not stack up well on either a total expense or adjusted expense basis. In fact, in only two categories--India Equity and Technology--did CEFs come out cheaper on either a total expense ratio or adjusted expense ratio basis. Both CEF categories proved cheaper on both bases. In the India Equity CEF category, there are two funds: India Fund
We can also look at equal-weighted category average comparisons (click here for results), and here the CEFs do a little bit better--but nothing to crow about. In 22 of 58 categories, the average CEF adjusted expense ratio was lower than the average mutual fund's. Two things jump out from this comparison. First, most of the categories where CEFs are less expensive are equity categories. Could this belie a belief among fund families that investors are willing to pay up for income but unwilling to pay up for long-term capital gains? Second, and no doubt related to the fact that they're primarily equity categories, the CEFs in the 22 less-expensive categories use little leverage. In fact, in 13 of the 22 categories, the CEFs don't use leverage. This suggests that, barring the use of leverage, CEFs could in fact be less expensive than open-end funds; in reality most CEFs do use leverage and incur the resulting costs.
Because of the difficulties in creating a fair comparison, there are many caveats to this study. Three are very important. First, these comparisons are based on one year's worth of data. Given that management fees are a large portion of expenses and that the management fee structure doesn't usually change for CEFs from year to year, it is unlikely that these comparisons would change much if we ran them again next autumn. Still, you never know. Second, all of these comparisons are based on averages. It is possible that a CEF you are considering may actually be cheaper than any other actively managed fund, and vice versa. Such details can get lost in large-scale, broad-swath studies like this one.
Sometimes, You Get What You Pay For
The third caveat to this study is that it only looks at expenses. In my mind, it would be lopsided to end the discussion here. Expenses (that is, costs) are one side of the coin, the other side being benefits. When it comes to actively managed mutual funds, Morningstar has done our fair share of pointing out that lower fees are the single best predictor of future total return performance. Does the same hold true for CEFs? Does it behoove investors to seek the CEF with the lowest expense ratio?
The answer is no. Leverage trumps expenses when it comes to returns. As long as a leveraged CEF's portfolio returns enough to offset the cost of leverage--and it doesn't have to be a raging bull market for that to happen when leverage costs are so low--then it pays off to pay up for leverage.
The table above returns to our three funds, used near the beginning of this article. Recall that funds 2 and 3 had leveraged the net assets up 50%. Here we assume that the portfolio total return for net assets in all three funds is 8.00%, and, by extension, the 50% leverage contributes an additional 4 percentage points to the total returns of the leveraged funds.
As can be seen in this example, the reported NAV total return for the leveraged funds is much higher than for unleveraged fund 1, despite the significantly higher expense ratios. Fund 2, with the cheaper cost of leverage, outperforms fund 1 by 2.87 percentage points, or by 41.5%, despite a total expense ratio that is more than double that of the unleveraged fund. Fund 3, with its high cost of leverage, similarly outperforms the unleveraged fund, even when we take into consideration the true economic cost of the leverage.
This fictitious representation of the benefits of leverage on a CEF's NAV total return shows that, in many cases and in good years, the extra costs associated with leverage--both from the higher management fees and the actual cost of leverage (interest expenses or preferred shareholder distributions)--can be more than offset by the benefits of leverage. Investors pay more for the funds, but they receive more in return.
How do category averages stack up when it comes to benefits? It's difficult to assess this, primarily for one reason. On the cost side, we looked at the most recent, one-year expense ratios. On the benefit side, it is rather aggressive to tout the benefits of a one-year return for a group of CEFs that are largely leveraged versus a group of mutual funds that are not leveraged (at least to the same degree and in the same manner). However, to line up the benefits over the past five years against the costs of the latest completed fiscal year would be disingenuous.
Still, we believe it is important to present the results over the past year. We are not backing off of our view that, to invest in leveraged CEFs, investors should have a three- to five-year investment horizon. Leveraged funds are volatile and thus represent true risk and require true patience to endure a market downturn. What we've seen over the past year, for most categories, is an environment where markets have performed well, leverage costs have been low, and conditions have favored leveraged funds. If we'd run this study in the fall of 2008, our conclusions would have been quite different.
Looking at asset-weighted total returns for the one-year period ending Oct. 8, 2012, CEFs outperformed in 37 of the 58 categories on a NAV basis (click here for results). What is striking is that, in 15 of the 22 categories where the CEF category equal-weighted average adjusted expense ratio is less expensive than its mutual fund counterpart, the CEFs had worse performance on a NAV basis. In 11 of the same 22 categories, the CEFs had worse performance on a share price basis. Furthermore, in the 36 categories where the average CEF was more expensive than the mutual fund average, the CEFs had better performance on an NAV basis in 30 of the categories and on a share price basis in 32. On an asset-weighted comparison, choosing CEFs on an expense ratio basis over the past year was not the right way to invest.
We see the same thing when looking at equal-weighted total returns (click here for results).For the 22 categories where the CEF category equal-weighted average adjusted expense ratio is lower, the CEFs underperformed on a NAV basis in 13 categories and on a share price basis in eight categories. For the 36 categories where that CEF expense ratio is higher, the CEFs outperformed on a NAV basis in 32 categories and on a share price basis in 33.
Now, again, these comparisons are only for a one-year period and in many cases pit leveraged CEFs against unleveraged mutual funds. Still, for CEFs, they show that low expenses do not necessarily lead to better absolute performance. Leverage--and its current low cost--trumps expense ratios when it comes to absolute, not risk-adjusted, total returns in an upward-trending market.
The data largely support the investor complaint that CEFs are more expensive than mutual funds, with a few category exceptions. However, understanding the sources of the higher expenses is important in order to discern whether there are benefits that come along with the higher costs. Leverage accounts for a large portion of the additional expenses incurred by CEF shareholders. At least over the past one-year period, the benefits of owning leveraged CEFs have outweighed the accompanying higher costs. When it comes to CEF investing, it is not enough to simply choose the CEF with the lowest costs. In fact, when the markets are cruising along and all else being equal, choosing the cheapest CEF is not necessarily a smart investment decision. Sometimes, it pays off to pay up.
Click here for data and commentary on individual closed-end funds.