Knowing the essentials of how the lendable value of loan collateral at FHLBank Boston is calculated can allow members to optimize their balance sheet and better manage liquidity risk across different environments. In the first of a two-part examination of this important determination, the essentials of calculating lendable value are explored.
Maximizing borrowing capacity may seem to be an academic exercise for most members, at most times. If a member generally borrows less or already has a reasonably significant amount in additional borrowing capacity, why should they bother trying to optimize finding further dry powder when there are so many competing and urgent priorities?
In short, there are two reasons to concentrate on borrowing capacity: contingent funding for unexpected stress events and leverage for moments of particular opportunity.
Recent high-profile bank failures demonstrate the need for the former. Indeed, the initial catalyst for the runs that ultimately brought down these institutions was, arguably (and ironically), overconfidence in asset-based liquidity. Because bond portfolio valuations declined on interest rate changes, large uninsured depositors lost confidence in these institutions and abruptly pulled their funds. While the size of the holes in the balance sheets proved too difficult to overcome in these particular cases because of the interconnectedness of the clientele involved, a less dramatically scaled or concentrated loss of depositor confidence at a different depository may have been able to be met with enhanced liability-based liquidity – that is, borrowing capacity.
Maintaining ample borrowing capacity could also help any type of member – whether a depository or an insurance company – take advantage of market opportunities. Spreads in fixed-income assets often widen during times of dislocation or confusion, and a shrewd institution may reasonably determine that the potential rewards at such moments justify the potential risks.
"Whether out of defensive prudence or offensive posturing, optimizing borrowing capacity based on an institution’s balance sheet is a key part of asset/liability management."
First, members must understand what their credit status category is because it impacts the manner in which loan collateral may be pledged. Three levels exist for depositories:
Category 1 members can potentially enjoy significantly lower haircuts if they opt to voluntarily list owner-occupied one- to four-family loan collateral instead of filing QCRs. For instance, on conventional one- to four-family residential loans, the minimum haircut is 18% when listed versus 25% when filed on a QCR. Insurance company members are not placed in the same credit status categories because they are subject to different state laws and regulations.
Second, there are overall governing limits that members should know whether they are depositories or insurance companies.
For the vast majority of members that list loan collateral, three major components – asset valuation, a haircut, and a collateral adjustment factor – are used to determine the lendable value of pledged loan collateral.
The first component of determining lendable value, the collateral adjustment factor (CAF) is derived from a member’s historical reliability in pledging eligible assets.
Periodically, underlying loan notes are reviewed by FHLBank Boston collateral teams. To determine the CAF, the number of loans pledged by a member found to be ineligible after the underlying documents and data are reviewed is divided by the total number of loans reviewed. For instance, if 50 loans were reviewed and one was found to be ineligible, the CAF would be 2% (1/50).
Second, the underlying value of the pledged asset is estimated. While this is typically straightforward for marketable securities, loan valuations are not quite as easily observed.
Two general sets of criteria are used to evaluate loans: 1) a discounted cash flow analysis that looks primarily to fixed-income dynamics and 2) the underlying property value that looks primarily to credit dynamics. Functionally, one or the other will likely dominate the valuation depending upon the economic environment. These valuations are made quarterly as of the second month of the quarter (February, May, August, and November).
Due to the combination of elevated property values and materially higher interest rates in the past year, cash flows are dominating valuation estimates. Like other fixed-income instruments, the rapid increase in rates has put downward pressure on fixed- and adjustable-rate loans with longer reset frequencies.
Beyond nominal interest rates, wider market spreads can also affect asset valuations. As with higher nominal rates, higher spreads also increase the discount rates used in present value calculations, which would result in lower valuations, all else equal.
At other times, however, declining property values may push the asset valuation lower based on market dynamics. If interest rates decline or stay the same since a loan was booked (increasing the discounted cash flow valuation), but the credit environment has deteriorated to the point that real estate prices decline significantly, the overall asset valuation could still be lower. Geography and property type are also contemplated in the credit-oriented valuation process.
While precise changes will be idiosyncratic to the particulars of pledged loans, in general, if interest rates are at or near multi-year highs in a declining real estate market, asset valuations are likely to be low. The inverse is true, too; if interest rates are at or near multi-year lows in a benign or strong real estate market, asset valuations are likely to be high.
Importantly, in no event can the asset valuation exceed the underlying book value of the loan. Even if the discounted cash flow analysis implied the equivalent of a bond price above 100 in a market with strong real estate price appreciation, the asset valuation would not exceed the nominal value of the loan.
Finally, a haircut is taken against the asset valuation.
Haircuts estimate the risk of decline in loan value given adverse movements in rates, spreads, and property values over some period of time. Essentially, the haircuts identify how much the asset valuation could deteriorate over an assumed holding period and in a liquidation event at a moment of significant market strain.
Haircuts are also specific to the precise characteristics of the collateral pledged, but, in recent calculations, haircuts for major category types are often in the following percentage point ranges:
To calculate the lendable value, multiply (100% - CAF) by the asset valuation by (100% - Haircut).
Based on a discounted cash flow analysis, a hypothetical residential mortgage booked at 3.00% in 2021 would have a lower asset valuation because currently prevailing interest rates are higher. Assume that the estimated value is 83% of book value. Assume that the underlying property value, however, is still equal to the purchase price, and the underwriting characteristics of the loan are relatively conservative, resulting in a haircut of 20%. The member’s CAF is 2% based on a recent review.
In this case, the lendable value of the loan would be calculated as follows:
98% post-CAF value x 83% asset valuation x 80% post-haircut value = 65% lendable value
Now that the lendable value calculation has been explained, we’ll examine ideas and methods for optimizing borrowing capacity based on these guidelines in the second part of this series. Please contact me at 617-292-9616 or sean.carraher@fhlbboston.com or reach out to your relationship manager for more details.
FHLBank Boston does not act as a financial advisor, and members should independently evaluate the suitability and risks of all advances.