The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
Ir1c;:;lffil Q) IT §@Jl'\l I F If@\ffil<r:: li Q) May 20,1 983 - - - _. _- - - _. _- . - - -- _. _- - - _. _- - - - Market Val u e- P art II Ou r Letter of last week described the concept of market value accounting and compared it to historical cost accounting for financial institutions. Market value accounting requires that assetsand liabilities be expressed in terms of their present values, i.e., the discounted-to-the-present valu.esof the cash flows they promise, using current estimates of interest rates expected to prevail over the life of the contract. In keeping with the present value concept, income in the current accounting period includes cash flows and realized capital gains/losses as conventionally measured with the addition of the current period's unrealized capital gains/losses that arise from changes in present values of assets and liabilities. In contrast, historical cost accounting requires that assetsand liabilities be expressed in terms of their contractual book values. Current income in this system includes only actual cash flows and realized capital gains/losses. Thus, historical cost is "blind" to projected cash flows beyond the current accounting period, while market value accounting includes the implications for present wealth arising from all future contractual cash flows. In this sense, it is superior to historical cost accounting. aggregate net worth of S&L's and mutual savings banks, although actual figures are subject to considerable margins of error. Among financial institutions, S&L's and mutual savings banks stand out by their practice of "lending long and borrowing short." Because lending long commonly involves fixed-rate contracts extending a number of years -I ike mortgages, for example-the present value ofS&L assetsis especially sensitive to changes in interest rates, or to what is called interest rate risk. The typically shorter maturity of their liabilities on the other hand impl ies less sensitivity to interest rates. As a result, the net worth of S&L's is particularly sensitive to swings in interest rates. (In contrast, because commercial bank business lending usually is confined to variable-rate and/or shortterm contracts, banks are better hedged than thrifts against interest-rate risk.) Estimates In looking beyond the present, however, market value accounting requires estimates of interest rates, prepayments, defau Its, and other factors that might affect future cash flows and their present values. These esti- • mates render market value accounting subject to some uncertainty, but nevertheless make it a useful adjunct to book value accounting. Calculating market value balance sheets for thrifts under different interest rate environments gives a measure of how market value net worth is affected by the level of interest rates. Perhapsthe most thorough study of the market value net worth of thrifts was conducted in 1981 by Richard Kopcke of the Federal Reserve Bank of Boston. Employing alternative assumptions for effective maturities of assetsand liabilities, Kopcke made estimates of the divergence between book and market value net worth for Massachusetts mutual savings banks and California S&L's from the late 1970s through mid1981 -a period of an unexpected and unprecedented rise in interest rates. In this Letter we turn to some of the problems of market value accounting and to market value estimates of the impact that interest rate swings have had on the net worth of thrifts. The boltom line is that interest rate changes have had large impacts on the Whereas aggregate book value net worth was approximately 8 percent and 6 percent of total assetsfor the respective thrift groups in mid-1 981 , Kopcke's calculations of aggregate market value net worth ranged between -1 percent and -1 2 percent for ()piniolls in this newslettpr do nClt P?f1ect thc' of the llldnagerrWIl'1 ()f the F'('deral Re'serve Bank ()-r Sdn L:ranris(:o, or oj the Board or' COVC'JTlOrS of tf"w Federal Reserve Svstem. Massachusetts mutual savings banks and between - 3 percent and -7 percent for Cal ifornia 5&L's. Perhaps as many as two-thirds of thrifts were estimated to have negative net worth at market value, that is, to be technically insolvent if interest rates were to remain atthe mid-1981 level! But it is important to appreciate the considerable uncertainty that is attached to these estimates. Alternative estimates of effective maturity show a considerable range. Data on 26- to 3D-year Federal Housing Administration (FH A) insured mortgages were compiled by H U D for the years 1 957-76. The probability of prepayment (including default) in a given year was estimated on the basis of actual experience, and this pattern for FHA mortgages has since been referred to as the "1 00 percent FHA experience." Mortgages following such a pattern would have an average life of 13.1 years. On the other hand, some experiments done by the authors using Government National Mortgage Association (GN M A) passthrough securities, which are backed by FHA mortgages, suggest a shorter effective maturity. These experiments compared the present value of the securities calculated under different assumptions about their maturity with their actual market price. Problems valuingassets in The assets in thrift portfolios consist mainly of fixed-rate mortgage loans. One of the critical issues in calculating the present value of a mortgage is its effective maturity. Mortgages usually are repaid prior to maturity, so that the effective maturity is less than the contracted life of the instrument. One can attemptto estimate the average lives of mortgages based on past experience but there is no reason for future prepayments to conform to past experience. Prepayment options, housing turnover, discrepancies between present and contracted mortgage rates, legal constraints on due-on-sale clauses, and defaults all affect the effective life of a mortgage and hence estimated present value. They suggest that investors priced the G N M A securities on the basis·of about a 9-year average life. Apparently, investors price such mortgage-backed securities as fairly short-term instruments, treating the mortgages that lie behind them as effectively medium-term rather than long-term securities. Because of this, high interest rates may have had a less drastic effect on the market values of thrift-held mortgages than is often suspected. Consequently, estimates like Kopcke's of negative net worth may have overstated the seriousness of the thrift situation. If all lending were contracted at interest rates that varied along with the level of openmarket rates, interest-rate risk due to large swings in rates would be reduced. Still, under current arrangements, they would usually not be eliminated since variable-rate lending often is only partially indexed. Contracted rates, especially in the case of variable-rate mortgages, move only partially 2 from -7 percent to - 3 percent. As this variation in estimates indicates, assumptions about the effective life of deposits are of critical importance. with open-market rates and often with prespecified lags and limits. Present values of such instruments are especially difficult to calculate in the framework of market value accounting. However, the expanding resale market for mortgage passthrough securities should provide a broader information base from which to determine their present values. True net worth Our analysis concludes that market value accounting must rely on inputs that are highly judgmental. Herein lies its weakness. Assumptions and imprecise calculations must somehow be made if one is to estimate the current net worth of financial institutions. Market value calculations are made implicitly everyday in the valuation of equities and are made explictly whenever mergers, acquisitions, or regulatory actions are contemplated. Historical cost accounting is adequate for showing book values and current cash flows, but it gives little indication of true net worth. Market value accounting offers the potential for coming closer to the truth, although the practical difficulties in making present value estimates in the current state of the art mean thatthese estimates should not be used uncritically. Valuing liabilities Determining the present values of liabilities raises even tougher issues, among which estimating their effective maturities is perhaps the most difficult. Many liabilities such as term fixed-rate deposits present relatively few conceptual problems. But even these deposits normally are subjectto early withdrawal at the option of the depositor (usually under penalty). Consequently, their effective maturity depends in part on future interest rate movements, economic conditions, and deposit alternatives. The more difficu It problem centers on determining the effective maturity of "core" (usually, "non-interest sensitive") deposits that can be withdrawn without penalty atthe depositor's option. Passbook savings for years have had a maturity of one day although these funds have not been withdrawn despite significant changes in openmarket interest rates. What figure does one use for their effective maturity or duration? Jack Beebe and Matthew Blank Kopcke found that passbooks, N O W accounts, and comparable core deposits made up 40 percent of Massachusetts savings bank liabilities and 16 percent of California S&L liabilities in 1981. Not knowing the effective maturity of these deposits, he assumed alternatives of one year and eight years. The corresponding variations in market-value net worth of Massachusetts savings banks varied from -1 2 percent to -1 percent, and that of California S&L's 3 SS'I110 J.Sl::lI::l UOlaU!4SI?M.41?1n • uo8aJO • epei\aN • oyepi H!?M!?H ,-. I?!UJoJ!II?) • I?UOZ!JV • 1?>j51?1V (\)) 'llIE::>'O::>SPU2J.:I ueS ZSl 'ON llWHld OIVd]9VlSOd 'S'O 11'11'1 ssvnlSHU :(l\Ill <OJ JJW!rdJ \Ijf;JJW!<OJ\B '!illl:(l <OJ@ JJ <OJQ:lj BANKING DATA-TWELFTHFEDERAL RESERVE DISTRICT (Dollar amountsin millions) Selected Assetsand liabilities large Commercial Banks Loans (gross, adjusted) and investments>!' Loans (gross, adjusted) - total# Commercialandindustrial Realestate loans to individuals Securities loans U.S. Treasury securities*" Other securities* Demanddeposits-total# Demand deposits- adjusted Savingsdeposits - totalt Time deposits - total# Individuals,part.& corp. (Large negotiable CD's) WeeklyAverages of Daily Figures Member BankReserve Position ExcessReserves(+ )/Deficiency (-) Borrowings Net free reserves(+ llNet borrowed(-) Amount Outstanding 5/4/83 164,141 142,094 45,541 56,251 23,631 2,787 8,097 13,949 41,562 28,305 65,626 66,354 59,537 19,670 Weekended 5/4/83 117 121 26 Change from 4/27/83 Change from year ago Dollar Percent 406 452 301 6 3 113 4 49 2,341 94 636 - 267 80 92 Weekended 4/27/83 4,501 3,286 2,278 908 223 662 2,068 852 1,686 1,565 34,796 26,103 23,389 14,250 2.8 2.4 5.3 - 1.6 1.0 31.2 34.3 - 5.8 4.2 5.9 112.9 - 28.2 - 28.2 - 42.0 Comparable year-ago period 72 135 63 55 17 38 '" Excludes trading account securities. # Includes items not shown separately. t Includes Money Market Deposit Accounts, $uperHNOW accounts, and N O W accounts. Editorial comments mayhe addressed the editor (GregoryTong) to the author.... Free to or copies of this and other Federal Reserve publications beobtainedby callingor writing the Public can Information Section,Federal Reserve Bank of SanFrancisco, P.O.Box7702,SanFrancisco 94120. Phone (415) 974-2246.