IFRS9 was conceived in the wake of the financial crisis to address criticism of the prevailing impairment model that allowed banks and other lenders to delay recognition of losses. The goal of the new accounting standard is for lenders to have more appropriate levels of credit loss provisions earlier in the credit cycle. Institutional investors welcome the accounting changes, particularly because they are intended to reflect the most current and complete expectation and estimation of the value of assets. Decision making and comparability about investment options are enhanced, institutional investors argue, when fair value is used to measure assets. We agree that a central consideration of the appropriate accounting model has to be the usefulness of the information to investor decision making. However, some of the financial assets held by private debt funds could actually deliver more “noise” under IFRS9 and potentially lead to flawed decisions.
It is not uncommon to see alternative lenders negotiate equity kickers and other incentives as part of their lending arrangements, often as trade-offs for lower contractual interest rates and fees. We usually obtain such supplementary benefits as consideration for the strategic consulting and operational improvements we provide borrowers. These assets are rarely reported on lenders’ financial statements (except as footnote disclosures) because they are received for no cost, are contingent in nature, and due to the lack of universally-accepted measures to reliably value them. Currently, it is our practice to recognize the value of financial derivatives of private companies, such as options and warrants, and contingent payments on future revenues, such as royalties, only to the extent that we collect cash on these assets or when these assets convert to financial instruments that are actively traded in liquid markets. IFRS9 changes the measurement of these assets and forces managers of private debt funds to ascribe them a fair value.
Effective January 1, 2018, fund managers must value options, warrants, and royalties at fair value with changes in fair value at each measurement date recognized in profit and loss as they arise (“FVTPL”). Prior to IFRS9, managers could rely on accommodations and exceptions to measure these private, non-traded derivatives and contingent payment assets at cost (usually zero) given the significant range of possible fair value estimates and probabilities. Measuring fair value of these assets is complex and potentially costly because it requires manager to consider various valuation techniques such as the market approach (recent transaction prices for identical or similar instruments), income approach (discounted cash flow), and adjusted net asset (market value of assets less market value of liabilities). A major consequence will be selection bias, where managers pick the model that provides the highest value. In addition, because of FVTPL, income statement volatility is bound to increase.
Managers can manipulate model inputs, making fair value estimates more subjective. Informational asymmetry and the potential for adverse selection combined with the moral hazard of having managers apply the information to fair value measurements in a neutral and unbiased way are issues that impact the reliability of such measures. The greater the subjectivity involved in a valuation, the greater potential for unreliability. Can fair value be reliably measurable for a long-dated call option on a private business? We do not think so as the range of possible outcomes is too large or irrational given the substantial time value. As Warren Buffett rightly points out, “If the Black-Scholes formula is applied to extended time periods, it can produce absurd results”.
Fair values are only relevant to the extent that are reliable. The use of manager discretion on calculating fair value has no unifying benchmark that can align assumptions across funds with any economic reality. IFRS9 potentially makes it easier for managers to downplay problems and evade an investor’s early warning signals by picking the right model to value an asset until, in the case of a private option for example, it expires worthless. Investors’ capital is needlessly put at risk when they depend on assumptions that cannot be relied upon.
Our views definitely put us in the minority. Most financial institutions, institutional investors, and auditors staunchly support IFRS9. They believe regulatory sanctions such as monetary penalties and investing bans for improper valuation are major deterrents that will compel managers to ensure a higher degree of representational faithfulness in their measurements of private assets. Regulators are not an effective disciplinary force to keep funds and managers honest because it does so with a significant lag. Proponents also point to research that shows investors have been able to, over time, see through attempts by managers of less healthy funds make their funds appear healthier by exercising discretion when estimating investment fair values. We do not see this taking place in practice. One notorious, publicly-traded asset-based lender has a market capitalization that exceeds its net loans receivables despite the fact that 78% of such loans are an “airball”, meaning they are not actually secured by any discrete assets and instead valued according to the lender’s models. The lender even, in some cases, used fair value estimates to justify marking-up an investment that it acquired through a credit bid from its own defaulted loan!
Investors should carefully consider how certain accounting rules under IFRS9 will impact the reliability of valuation estimates provided by their managers. To assess reliability, investors should request transparency into all of the elements that make up a valuation for private assets determined by a model. For private debt investors, good governance and consistently-applied processes that are well-defined will be hallmarks for reliable valuations. The potential for large profits or losses from re-marks is enormous. In periods of rising asset prices, FVTPL will accelerate the recognition of gains, and therefore the compensation of managers. The question of whether IFRS9 will provide intended benefits in a downturn through fewer unexpected markdowns will eventually get answered.