Back in 2005, the College Board revised the SATs so that a score of 1,600 no longer represented absolute perfection. After 70Ys, students readjusted to a maximum score of 2,400. While the infamous analogies section was removed, essay, algebra and new grammar segments were added. This month, SAT test-takers across the country will attempt to find out what the new "benchmark score" is. For 2007, the average SAT in Math was 515 out of 800, the average score on the Critical Reading section (which used to be called the Verbal) was 502 and the score on the new Writing section was 494. It is rumored that for 2008, scores of 760, 730 and 740 will be needed to get into Harvard.
As can be seen above, benchmark scoring is helpful, as it gauges some of the capabilities of a student. The benchmark can also provide valuable information about strengths and weaknesses. In like fashion, many larger banks turn to the Libor or Swap curve to benchmark themselves when it comes to loans, investments and liabilities. By maturity structure, large banks use Libor rates to benchmark things in the first 12 months and then switch over to the swaps curve out to 30Ys. In so doing, these banks receive a readily available and liquid index, against which they can gauge how well their loan pricing, investment returns or cost of funds stack up and change over time. A 1Y cost of funds priced at Libor less 50bp last year, compared to Libor less 75bp this year, is an improvement of 25bp and shows that management is proactively restructuring their liabilities. Some banks will also benchmark against FHLB Advance rates. This is fine, but it has the disadvantage of being a less liquid measurement stick, is prone to specific supply and demand considerations not found in Libor (the most liquid market in the world) and can be different depending on the FHLB district. Treasury indexing is another choice utilized occasionally by bankers, but it has the disadvantage of dislocating when markets get choppy, caused by high demand for certain issues or due to low supply in certain maturities (like the 30Y).
For banks, benchmarking has the advantage of helping with asset-liability management, as banks can run correlations to judge what index best approximates their cost of funds. This can help when analyzing how to price a loan, hedge it or project future funding costs based on market expectations of Libor. For example, a typical community bank has a 92% correlation to 1-month Libor for funding. Said another way, 92% of the time, a bank's funding cost will closely track 1-month Libor in some consistent and calculable fashion. Oftentimes, banks will lag their index to get an even closer correlation. This recognizes that the bank will take anywhere from 3 to 12 months to adjust pricing in a changing rate environment, because of their deposit structure. Once adjusted for this lag, correlations can near 98%.
In addition to providing greater insight into performance and risk management, understanding how assets and liabilities react to the market and each other is important. This allows banks to build simple funds transfer pricing that can then be more closely tied to customer and product profitability. Recent market volatility, combined with the quick drop in the Fed Funds rate has given banks a rare opportunity to reexamine how prepayments, deposit duration, liability run off and general performance are correlated to sharp drops in interest rates. Up to this point, it was largely a theoretical argument about how movements in rates impacted performance. Bankers now have live examples to verify how their bank has reacted to this extreme environment and adjust their assumptions for more accurate future modeling.
While bank CFOs luckily don't have to sit for the 4 hours of misery the SAT provides, incorporating benchmarks into bank management will have the advantage of producing a more consistent way to track performance over time. Through benchmarking, maybe one day CFOs can join the 107 students (out of more than 300k) that achieved a perfect score.