A reader requested comments on the paper “Why Do Closed-End Bond Funds Exist?” by Edwin Elton, Martin Gruber, Christopher Blake and Or Shachar. This study investigates the unique aspects of closed-end bond funds using characteristics and performance data mostly from 1996-2006 for two samples: (1) 54 pairs of closed-end and open-end bond funds matched for manager, fund family and type of bond fund; and, (2) 332 closed-end bond funds. The essence of their findings (from the “Conclusions” section of the paper) is:
“…almost all closed-end funds borrow [to lever their portfolios], where open-end funds do not. The borrowing is short-term while investment is long-term. We find that closed-end funds do not vary their borrowing as the term premium changes, but they do time issuance [mostly of auction rate preferred stock] to coincide with a high term premium [steep yield curve]. Because they do not vary their borrowing, the return they earn is directly related to the spread between short term and long term rates.
“…leverage increases returns in most years and decreases the correlation of returns with security market indexes, but it also increases the variability of return. The net effect of these three influences leads to investors being better off. We show this in three ways. First, levered funds have higher alphas than unlevered funds and matched open-end funds computed with respect to either a bond index or a stock index. Second, using mean-variance analysis, levered funds are more likely to enter a bond or stock portfolio than unlevered or matched open-end funds. Third, investors pay more for a dollar of net assets (smaller discount) in levered funds than in unlevered funds. Leverage explains much of the cross-sectional and time-series variation in the closed-end funds discount.
“…even after the credit crisis closed-end bond funds have continued to outperform their open-end bond fund counterparts.”
In summary, because of leverage, closed-end bond funds on average produced higher alpha and better diversification relative to stocks and bonds than open-end bond funds over the 11-year sample period.
Things to wonder about are:
- Is the 1996-2006 sample period that serves as the focus of the study long enough in terms of number of interest rate cycles to form a confident expectation of future closed-end bond fund performance?
- Is the market changing in ways that will depress bond returns and/or reduce the value of the leverage used by the closed end funds (because of reduced spread between bond yields and borrowing costs)?
Regarding the latter point, the study states:
“…the failure of the auction market will reduce their advantage as an investment vehicle. Alternative sources that are being used are more costly. Only time will tell how large these added costs will be. However, given the historic spread between return on assets and borrowing costs, we do not expect these added costs to eliminate the advantage of leverage for closed-end bond funds.”
