My recently published article on the correlations between leverage and returns for closed end funds (here) was met with skepticism from some readers. Among the objections was the suggestion that I may have been over-interpreting results from scanty data. Another involved the validity of lumping all closed end funds together. And a third concerned the limited time duration considered.
I accept that each of these has validity and present here some additional data to partially meet these objections. I say partially because given the limitations of publicly available data sets for closed end funds it is not practical to address those objections in a way that would pass a rigorous statistical review. So, the results here should not be taken as definitive. I submit, however, that they are suggestive, sufficiently so to call into question the received wisdom that increasing leverage will enhance returns in an up market. I will show that this is clearly not always the case across the spectrum of closed end funds.
Keep in mind that closed end funds are all actively managed. These are not passive funds following an index. Many investors are familiar with leveraged ETPs that have unleveraged counterparts following the same passive index. In such cases, the two products share the same asset portfolio. One reasonably expects -- and the data support this expectation -- that returns will be multiplied by the leverage factor. Thus a 2x leveraged product will outperform its unlevered counterpart by 2x (less fees) for positive returns and underperform by that same 2x (less fees) for negative returns.
By contrast, closed end funds are actively managed. Actively managed funds do not share the identical asset bases. For closed end funds managers have a lot of flexibility, so within categories they may employ markedly different strategies to obtain results. Thus the leverage multipliers will only apply to the results of a given management team.
Another consideration is that many of the leveraged index EFTs and EFNs are sponsored by entities that have access to extremely low cost borrowing. This is often not the case for CEFs, which typically have higher borrowing costs which tends to reduce the gains (and exaggerate losses) that might be otherwise expected from simply applying a multiplier based on effective leverage percentages.
What I will present here is a more granular set of correlations between effective leverage and fund returns by category than presented previously. In the last analysis I limited myself to categories with at least ten leveraged funds. I drop that down to seven funds this time around to allow more categories. This tends to exaggerate the impact of extreme underperforming or overperforming funds, particularly at the edges of the range of leverages, but it does permit analysis of a greater number of categories.
I also look at one and three year returns for both market price and NAV (net asset value per share). NAV is a better indicator of how well or poorly managers are performing. Market price filters that performance through market perceptions of the fund's performance as well as the broader perception of the asset class. Of course, it's market returns that affect one's balance sheet.
I limit myself primarily to equity funds here. The reason is that equities have enjoyed a steady bull market for the one-year and 3-year periods I consider; fixed-income asset classes have not. If the conventional wisdom is valid, leveraged funds should produce greater returns at rates approximating the level of leverage in such market environments. Of course, not all equity asset classes have risen at the same pace; some have seen periods of decline. But in no case has any of the funds in the included categories fallen into negative territory for the one- or three-year periods considered.
Domestic equities were the biggest surprise for me and for readers. Surely in this bull market for stocks, leveraged funds should outperform and funds with greater levels of leverage should show commensurate levels of outperformance. The results show that this simply has not been the case.
US Equities: General Equity
Looking first at the domestic general equity category we see that one-year market and NAV shows a negative correlation with leverage. And, unlike most of the more broadly defined categories examined previously, the correlations are fairly strong with r2 values of 0.341 for market price and 0.306 for NAV.
Three year returns show a slight positive trend for market price, but a slight negative trend for NAV returns.
I've added median values for the unleveraged funds from the US Equity-Covered Call category to these charts for a comparison. Note how market returns for the unleveraged equity funds compare quite favorably, especially to the more highly leveraged general equity funds, for the past 12 months. Interestingly, this is less the case for NAV returns where median returns correspond more closely to returns from the lower performing leveraged funds. This reflects the reduction in the market discount for the covered-call fund category for one-year. It seems there has been a move into covered call funds, likely driven by cautious investors concerned about an imminent correction in equities as covered call funds are considered more strongly defensive. For three years, the comparison between the leveraged and unleveraged categories is closer to what one might have predicted: During that bull market period the leveraged fund have generated stronger returns.
US Equities: Tax Advantaged Funds
Next up is the CEF category of tax-advantaged domestic equity funds. These funds are managed to provide tax-advantaged income. Managers use various strategies including options, qualified dividends and strategic selling of holdings to generate tax-advantaged returns.
For one year there is a slightly negative correlation with percent effective leverage. This despite the fact that for the year returns were positive greater than 20% on NAV across the category. For three-year returns funds in this category perform more like one might have predicted, although these correlations are driven disproportionately by one underperforming fund at the low end of the leverage scale. This is particularly the case for the market price return correlation where that underperforming fund generated less than a third of the gains of the rest of the category. Regardless, the r2 values indicated that meaningful shares of the positive trend (27% at market price and 43% at NAV) can be accounted for by the percentage of leverage employed.
I did not include the covered call funds for an unleveraged comparison here, but recall that the one-year values were near 20% and the three-year values were near 10%. Thus, this group of leveraged funds has generally outperformed that unlevered group. Whether this is a consequence of leverage or simply a reflection of the two different strategies is an open question.
US Equities: Growth and Income
The next category of domestic equity funds is equity growth and income. These funds have quite variable portfolios and strategies. All contain domestic equity holdings, frequently heavily invested in dividend stocks. The funds also generate income from options strategies and from holding various classes of income securities.
What we see here is, to some extent, the inverse of the previous category. One-year returns are positively correlated with leverage for NAV returns. Market price return flat. Three-year returns are strongly negative. And once again, the correlations are strongly driven by a single anomalous performer at the edge of the leverage range. The difference is that this time the outlier is an outperformer at the low end, thus it drives the correlation negative. One can in this case legitimately remove that outlier from the analysis. If so the three-year results are again flat with no benefit to returns attributable to leverage over the full range.
Utilities are well represented among closed end funds largely because they are the most traditional income-producing sector for equities. The results for the US equity utilities category follow.
One-year results seem to follow the intuitive pattern: Results correlate positively with leverage for both market price and NAV. Three year results, however are, once again, flat with no apparent effects of increasing leverage.
Preferred stocks is another income category well represented in the world of closed-end funds. It is also an arena where leverage figures heavily. All of the preferred stock funds are leveraged, generally near the top of the range for leveraged funds. There is, therefore, no unleveraged comparison fund and the range of leverage is fairly tight making the kind on analysis I'm doing here less than informative. With that consideration in mind, let's look at the data:
Once again we find at best little more than trivial correlations between leverage and returns.
One final category of domestic equities that has at least seven leveraged funds is real estate. Here, at last, we have a category where the expected correlation between leverage and performance seems to hold with some consistency.
Increasing leverage is well correlated with increased returns for market price and NAV over one or three year periods. The effect of leverage is particularly well correlated for the 3-year period. R2 values are greater than 0.5 here indicating that leverage is accounting for more than 50% of the trend to increased returns.
The real estate category is not exclusively an equity category. Several of these funds venture into real estate debt including mortgage-backed securities, other real estate debt securities and REIT preferred stocks.
Global Growth and Income
Finally here's a look at a non-US category: global growth and income.
The now-familiar pattern of essentially no correlative effect from leverage is seen once again. The correlation is negative for one year and flat over three years. Although returns are positive, the periods under consideration have not been strong for global equities, so a factor to consider here is that the more highly levered funds may be experiencing enhancement of downside movements. In any case, in the absence of the strong bull market that characterized US equities it is difficult to make too much of the poor return obtained from leverage here.
These analyses paint a somewhat different picture from my previous article. In some of these categories, there does appear to be a gain to be had from applying increased leverage. Real estate most strongly benefited from leverage. But, overall, one finds little to support the notion that leverage is a desirable feature among equity closed-end funds even in the recent bull market conditions.
The most unexpected results involve domestic equity. For the trailing twelve months, a period characterized by a strong upward trending market, unleveraged funds (from the category of covered-call funds) performed at least as well as the leveraged general equity category. In other categories this is less strongly the case, but in none does the benefit from high leverage appear to outweigh the clear risk such a strategy entails. It would be particularly informative to have data from less bullish times to fully understand the impact of leverage risk here.
Overall, these results emphasize the importance of carefully considering how heavily leveraged a closed end fund is. The prudent investor would be well advised to consider carefully before investing in the most highly leveraged closed end funds.
Disclosure: The author has no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article. I hold positions in several equity closed end funds, all from the covered-call and tax-advantaged categories.