Our last post used a comparison of loan commitments, which re-proposed Rule 18f-4 would treat as “unfunded commitment agreement,” and “to be announced” (“TBA”) mortgage-backed securities trades and put options for bonds, which Rule 18f-4 would treat as “derivative transactions,” to isolate features that could be used to delineate these commitments based on their leveraging effects. Our next post will continue this analysis by considering whether each unique feature of a loan commitment mitigates its potential “leveraging effects.” Before proceeding, however, we will consider one feature shared by loan commitments and TBAs to illustrate why even seemingly common features may have different leveraging effects.


Loan commitments typically permit borrowers to prepay their loans at any time without penalty (although there may be breakage costs). This creates uncertainty as to the term of the loan, which makes it less likely that someone would pay a premium for the loan. In a world with positive interest rates, why would you be inclined to pay $101 for a loan that may repay only $100 and a penny’s worth of interest tomorrow?

Investors should be even less likely to pay a premium to acquire the loan commitment, as they cannot know when loans will be drawn and for what amounts, adding to the uncertainty as to cash flows. If the loan commitment is unlikely to have a positive value, it cannot increase a fund’s net asset value and, thereby, increase the speculative character of its shares, which is the “leveraging effect” that Rule 18f-4 would limit.

Borrowers may also prepay their mortgages at any time without penalty, which would also deter someone from paying a substantial premium for an individual mortgage. A TBA, however, is a commitment to buy an interest in a pool of over ten-thousand mortgages from across the country. These borrowers are unlikely to repay their mortgages all at once; in fact, investors may use statistical methods to model the rate of prepayments under different market conditions. In addition, agencies issue MBS on a regular schedule, so there is no uncertainty as to whether and when the commitment must be performed.

Consequently, investors will pay premiums (and offer discounts) for TBAs with coupons above (or below) current mortgage rates. Resulting fluctuations in the market value of TBAs can produce corresponding fluctuations in a fund’s share price. Therefore, even though a commitment to make fixed rate loans and TBAs may share common investment risks, greater uncertainty as to when and to what extent a loan commitment will be drawn and repaid should make it less volatile than a TBA. This is one reason why the leveraging effects of a TBA should be more pronounced than the leveraging effects (if any) of a loan commitment.


If a loan commitment does not fluctuate in value, then it cannot “present an opportunity for the fund to realize gains or losses between the date of the fund’s commitment and its subsequent investment ….” We regard this as the chief reason that Rule 18f-4 should impose fewer restrictions on “unfunded commitment agreements” than on “derivatives transactions.” Nevertheless, we would not propose to use prepayments as a criterion when defining an “unfunded commitment agreement.” First, because there is no corresponding feature of equity commitments. Second, because it would require inquiry into the prepayment terms of the loans, which can be quite technical. Instead, we will examine the other unique features of loan commitments in hopes that one or more of them will provide a practical means of identifying commitments without leveraging effects.

*     *     *

Please subscribe to our blog to stay updated on COVID-19-related and other asset management industry developments.