Allowance for Loan Losses Brings Planets Together

Regulators are on Venus, credit unions are on Mars.


For the past two years, I’ve been lecturing and consulting with my clients on various matters pertaining to the Allowance for Loan Losses (ALL).  The issues often include outdated methodology and have frequently had a critical impact on the continuation of the credit union as a going concern. If an outsider were privy to these consultations, they might conclude credit unions and regulators (and their auditor cousins) have been living on different planets, with differing interpretations of GAAP requirements and differing views of the credit quality measurements on planet Earth. 

In the past, I have written about basic ALL accounting requirements. As the recession ends and we begin to see the proverbial light at the end of the tunnel, I thought it would be helpful to address the common problems of the past and begin a conversation about what the ALL trends should reflect on a go-forward basis if in fact the worst is behind us.

1. Failure of Credit Unions to Own the ALL Equates to a Regulator Technical Knock Out (TKO)

You might have a reasonable ALL balance, but absent a thought-out and well-documented policy and supporting documentation, you can’t play the game. Many credit unions were late in increasing their ALL in late 2008 and 2009 because they did not have a smart approach. Don’t let the same problem make you late to decrease the funding of the ALL as the economy starts to recover from this recession. More on this in item No. 4 below.

2. Loss in Underlying Collateral Value Does NOT Necessarily Equate to Impairment

I’ve been working with many Sand State credit unions with large concentrations of residential mortgage loans that have incurred collateral devaluations in the 50% range. If you have a portfolio of mortgage loans with these characteristics, assessing impairment can be difficult. I’ve seen regulators and auditors erroneously conclude that collateral value declines are an automatic indicator of impairment. On the other hand, I’ve seen many credit unions ignore steep collateral declines and continue to simply use historical loss ratios to assess impairment. Venus meets Mars!  Collateral value can’t be ignored. 

Best practice would provide for much deeper analysis of the components of the portfolio, such as credit score diminution, length of ownership of property, and degree of collateral decline. Even the “worst of the worst” loans will not all default. There is no easy answer to the question of providing for losses on reduced collateral value loans. My personal skepticism is maximized when I see simplistic solutions to this complex question.

3. TDR Accounting Requirements — Nothing New Here

The GAAP requirements for TDR accounting are not new. If you have been criticized for inappropriate TDR accounting, there is a good chance your ALL policy and methodology were not well thought out. Although TDR accounting may involve significant and ongoing present value calculations, there are practical short-cuts that can be employed to minimize the bookkeeping headaches. Don’t give away the TKO. Tackle the TDR accounting requirements proactively — you’ll   score some good points with your regulators and auditors.

4. Expectation for Reduced PLL and ALL

The accounting standards introduced a concept referred to as “Directional Consistency.” This means there is an expectation that during periods of worsening economic trends, PLL will be increasing, and during improving economic trends, PLL will be decreasing. 

The Sand States have experienced pronounced negative trends over the past few years. Many of my clients tell me (and evidence is starting to support this) real estate values have firmed, employment trends are improving (or not getting worse), and delinquency rates are on the mend.  Further, many of these clients obtain independent third-party analytics of property values that indicate noted improvement from the lows of 2009. Assume these attributes are accurate for illustration purposes. My expectation would be that the amount of ALL funding should be decreasing, and in fact it might be appropriate to begin a gradual reduction of the ALL balance.

There are many critical factors that need to be considered before implementing an ALL reduction strategy. At some point in the recovery, this WILL happen (100% guaranteed). It’s time to start evaluating the timing of this critical question.

5. Board of Director Governance and the ALL

The ALL is the only significant account in the financial reporting structure that requires significant judgment and can result in significant financial statement error. Far too many credit unions have not engaged in on-going education with the Board of Directors on ALL methodology. In my opinion, this should be a quarterly conversation, and there should be a formal annual review of the ALL policy by the Board of Directors. 

Finally, remember the ALL is not supposed to be a measure of the fair value of the loan portfolio. Just because an investor might only be willing to pay 50% of book value for a loan or group of loans, the ALL would not necessarily be (probably not be)  50% of the loan balance. Why? Because the investor is making assumptions about the ultimate collectability of principal and interest based on current and future expected impairment characteristics. Further, the investor is also considering a risk-based rate of return, and the coupon rate on the loan is not relevant to the investor’s fair value calculation. However, for GAAP purposes, the market interest rate is ignored for ALL purposes. Further, assumptions about default caused by future events are also not relevant to the ALL calculation. Understanding the theory, building strong policies and analytics, and increasing Director governance are key to a successful ALL outcome.




Sept. 21, 2010


  • would these same general comments apply to the allowance account for the NCUSIF? NCUA is stating that they are following GAAP--what might differ from the allowance requirements you layout above?
  • enjoyed
  • Very informative. It is always gratifying to know that there is someone of Mike's caliber that has not only the knowledge of ALL Accounting, but a pulse on the industry. With the FASB's proposed, and very complicated, impairment model on the warming tray, he will be in high demand.
    Perer Sainato
  • Lots of good information in a concise format. Each statement can be an article in and of itself...Thank you...from an examiner from Venus.
    Emilie Kroen
  • In all of the due diligence we have conducted as part of merger negotiations in over dozen cases, we have seen that the allowance is in every case understated.

    The common mistake is that credit unions use historical loss rates to determine future losses. That does not work. Credit unions often fail to record the original credit scores for their loans and then do not track changes in credit scores over the life of the loan. We suggest quarterly refreshes of credit scores during periods of high unemployment and financial distress. Credit Unions also need to reassess property values with automated third party assessments.

    Future loan losses are most likely to directly track changes in the unemployment and the price per square foot of homes that are sold.

    I do not fully subscribe to the notion that if property values are impaired that there is not impairment in the loan. I would recommend that any loan with impaired property value and impaired borrower's credit score should require a member contact to talk about the borrowers ability and intent to continue making loan payments. After talking with the member an informed decision can be made with regard to adequately reserving for the loan.

    One of the biggest risks is that borrowers with the ability to repay their loan are walking away because they don't beleive the home value will ever increase. If the loan amount exceeds the home value by a signficant amount there is a huge risk of loss even if the borrower has the means to repay the loan.

    The best practice for credit unions is proactive member contact to help members before the member's problems become too difficult to resolve.

    This crisis is different than the usual recession. I believe we have another five years of high unemployment and soft or falling housing prices. I think it is far too early to think about reducing the allowance for loan losses. Our credit union has one of the first to increase the allowance ahead of this crisis and we will be one of the last to lower the allowance. The auditors where wrong in 2004 to 2006 by insisting that excessive allowances be trimmed. They are likley to be wrong if they suggest reducing the allowance given the current economy. Be strong and resist the urge to reduce the allowance.

    I see no indications that lowering the allowance makes sense with unemployment at double digit levels and over 40% of all home sales are of distressed properties that are selling at 20% or more discount to normal prices. There is ample evidence that their is a big backlog of properties yet to come on the market and there is no evident demand for these properties.

    The reality is that we have lost a lot of jobs and we have not yet become to replace those jobs and are not likely to do so for many years. Unemployment and low home prices are the key to our loan losses. Neither unemployment or low home prices show any sign of changing in the near future in any significant way.

    henry wirz