What is wrong with orthodox credit ratios?

These ratios fall short for three reasons.

• First, these ratios are static snapshots in time.

Even using the second order derivatives (i.e., the sequential change in these ratios) fails to provide a true insight into the underlying formation of credit risk at a bank. The orthodox ratios are quite malleable to bank actions depending on the speed and manner with which they manage non-performing assets. The timing of charge-offs, initial valuation of assets received in lieu of payment, which regularly resulted subsequent losses on previously-booked recoveries, reliance on delinquent loan sales and any subsequent losses, for example, mask the underlying true flow of new non-performing assets in a period.

• Second, these ratios are not comprehensive.

They only use the official reported non-accruals and not the various other categories (depending on jurisdiction) that manifest deterioration, such as long past due but still accruing, restructured, in foreclosure, other real estate owned, assets received in lieu of payment, etc.

• Third, these ratios, even the Texas ratio, do not accurately reflect the risk that credit poses to earnings or to future dividends or buybacks or even capital raises.

What is the “Cybiont CScore” and its associated series of ratios?

Put simply, the “Cybiont C Score” is the current stock of all classes of under-performing assets plus the expected level of newly created under-performing assets over the next 12 months (using the most recent period’s formation rate) divided by the available resources to confront these under-performing loans, namely existing credit reserves plus real tangible equity plus expected NTM pre-tax pre-provision earnings. This ratio is a probabilistic indicator, which as it rises, correlates initially with negative earnings evolution through higher provision and other credit-management-related expenses, and as it progresses higher, to capital actions (i.e., cancellation/reduction of dividends/buybacks or capital raises) or even failure. Variants of this ratio would be to include in the denominator only core pre-tax pre-provision earnings (i.e., exclude volatile earning streams such as trading), excess loan loss reserves above some minimum and excess regulatory capital above some minimum.

In addition to the “Cybiont CScore”, there is a series of other useful associated ratios that we use. For example, the adjusted non-performing loan ratio (the stock of all forms of under-performing assets divided by loans plus the other non-performing assets), the annualized formation rate (the annualized sum of the period change in adjusted non-performing assets plus period charge-offs plus period losses on other real estate owned(OREO)/assets rec’d in lieu of payments plus period losses on delinquent loan sales all divided by the previous period closing balance of performing loans), and the adjusted charge-off ratio (period charge-offs plus period losses on OREO/assets rec’d in lieu of payments plus period losses on delinquent loan sales all divided by the previous period closing balance of loans).

Latin origins

My sell-side research career began in the mid 1990’s in Mexico City providing equity research coverage of the Mexican Banks for the clients of a Spanish Investment Bank. Mexico, having privatized its formerly nationalized banking sector, along with the rest of Latin America, was the land of plenty for US and European investment banks, who rushed to set up shop, hang their shingle out and offer investment advice on the regions companies to their clients.

With a wave of privatization and liberalization rolling through the major sectors across the region, the opportunity was so palpable for these investment banks, and the dearth of experienced analysts so pronounced, that I soon realized after I started that almost universally, my Latin American colleagues and competitors alike, had little tangential, if any, previous sectoral, Latin American, or even any relevant analytical experience prior to being hired.

My first boss’ career path illustrates this point. A little more than 5 years earlier than my hire, being a Mexican with a post-grad philosophy degree and English fluency was sufficient to be hired as an entry level support analyst for Latin American FIG investment banking at a white shoe U.S. investment banking house, even with no analytical or sector experience. Not long after, poached by a legendary but now defunct European firm to be their sell-side Mexican Bank equity research analyst, two years, a parental leave and another headhunted move later, and my boss is the Head of Latin America Bank Equity research.

I, on the other hand, had 5 years direct industry experience prior to my hiring, starting with branch management & commercial credit lending roles at a Canadian bank before my MBA then strategic and acquisition valuation roles post MBA that culminated with several bank acquisitions in Latin America. Tasked with covering the listed Mexico banks as a sell-side equity analyst, I approached it in the same manner as I did for potential target acquisitions for my previous employer. I relay this background to explain why I came to develop a series of ratios, one of which later colleagues would nickname the “L-Score” and what we now refer to as the “Cybiont C Score”.

In those days investors were lured to the growth in the region. However, the critical characteristic to comprehend when investing in Latin America Banks—coverage of which I took over from my boss a year into my hire—was risk. Apart from Macro/Country risk, understanding and accurately forecasting credit risk was critical. So, for me, with my industry acquisition experience, I found the orthodox credit ratios, such as Non-Accrual, Charge-off, and Texas ratios relied upon by investors and analysts to be wanting.