Before the financial crisis, closed-end bond funds could access relatively cheap short-term funding through auction-rate preferred securities (ARS). This funding became expensive during the crisis when the ARS market collapsed. Costlier funding and negative bond returns during the crisis led to the funds, most of which were near their allowed leverage maximum, delivering to stay within regulatory limits.
Closed-end funds are traded on exchanges, and the number of shares is fixed at the outset. Hence, when an investor wishes to redeem his shares, they are sold in the market at the ruling market price; the closed-end fund does not have to sell securities to meet this redemption. In a perfect setting, the market value of the shares of a closed-end fund should be equal to the market value of its underlying securities, but in practice, closed-end funds usually trade at a discount to their net asset value. The authors explore how leverage affects the expected return and risk of a closed-end bond fund.
How Is This Research Useful to Practitioners?
Until the early stages of the financial crisis, levered closed-end bond funds generally borrowed short term via auction-rate preferred securities (ARS), which are long-term floating-rate securities whose interest rates are reset on a periodic basis using Dutch auctions. As of December 2007, $64 billion in ARS had been issued by closed-end funds. Under ARS terms, a closed-end fund using this funding has to pay a predetermined interest rate if the auction fails. Consequently, ARS funding became very expensive during the financial crisis, when the ARS market collapsed. With the drop in bond values, funds had to deliver to stay within regulatory limits. Closed-end funds redeemed their ARS by either using alternative financing methods or selling securities.
The shock to the funding of closed-end funds led to an opportunity to explore the connections between funding liquidity risk and market prices. Increasing leverage in a closed-end bond fund increases both its expected return and its standard deviation but does not increase its Sharpe ratio.
The authors demonstrate that the debt of closed-end bond funds has little credit risk and is, in fact, virtually riskless, even during the depths of the financial crisis, for reasonable levels of leverage. The cost of funding via ARS is disconnected from debt risk because it is affected by such factors as funding liquidity risk rather than bond fundamentals, which has caused problems for the closed-end bond fund industry.
When a fund delevers by selling assets, there is usually a transfer between equityholders and debtholders. With a predetermined desired leverage ratio to adjust to in the process of deleveraging, it is possible to find the amount of assets to be sold (whose proceeds are used to retire debt) for there to be no wealth transfer between stockholders and bondholders.
How Did the Authors Conduct This Research?
The authors model the net asset value and debt of a closed-end bond fund using the standard Merton model to analyze the expected return and standard deviation of the two components. In the case of an unlevered closed-end fund, the net asset value is viewed as the sum of its constituent bonds. In the case of a levered closed-end fund, the net asset value is viewed as a call option on a portfolio of corporate bonds and the debt is viewed as the sum of a long position in a risk-free bond and a short put option on a portfolio of risky bonds.
Closed-end funds are restricted by the Investment Company Act of 1940; the amount of debt cannot exceed one-half of the fund’s net asset value. The numerical example the authors use is of five corporate bonds, where each bond has a five-year term to maturity, the asset volatility of each firm is assumed to be 20%, and the constant risk-free rate is 5% annually.
In the case of an unlevered fund, higher correlation between assets leads to a larger return standard deviation (and vice versa); the expected return remains unchanged. In the case of a levered fund, increasing leverage provides no improvement in the Sharpe ratio and the debt of the levered closed-end fund is riskless for practical leverage levels.
The authors next examine the impact of the crisis on leverage and find that if the funds’ debt had been funded using long-term bonds rather than ARS, the impact of the crisis on closed-end bond funds would have been much less severe.
The authors provide insight by sifting through closed-end bond funds that persevered through the financial crisis to determine whether they survived because the funds had a more long-term stable funding source rather than because they generated enhanced performance through active management of bond assets.