US Banks: Margin Nirvana – Rising Short Rates and a Steepening Curve

ginna
Print Friendly, PDF & Email

SEE LAST PAGE OF THIS REPORT Howard Mason

FOR IMPORTANT DISCLOSURES 203.901.1635

hmason@ssrllc.com

September 5, 2013

US Banks: Margin NirvanaRising Short Rates and a Steepening Curve

  • Forward rates have become more favorable to bank margins since we published our note on July 13th calling for bank net interest margins to increase to 4% by mid-2015 from 3.2% today (on average across FDIC-reporting banks). Since net interest income accounts for approximately one-half of bank revenues, this 25% increase in margins is significant.
  • Despite more favorable forward markets, we are not raising our forecast for bank net interest margins but not there is higher margin for error against possible offsets such as a compression in loan spreads.
    • To entirely wipe out the margin benefits suggested by the forward rate markets, loan spreads would need to fall to the multi-decade lows of 2005-2007 when leveraged competition from the financing markets (CMBS and CLOs) was at is most intense; financing markets are recovering (see Exhibit), but the competition is nonetheless more muted.
  • It is possible we may repeat history from 2003 to 2006 when short rates increased meaningfully yet bank net interest margins were flat to down. However, from 2003 to 2006, the benefit from rising short rates was offset by a flattening curve; today’s forward markets, in contrast, are suggesting short rates will rise and the curve will steepen. This is margin nirvana.
  • Of course, outcome interest rates will probably not follow the forward markets but this risk to the analysis can, by the definition of forward markets, be hedged at no cost (beyond the possible need to post collateral).

Exhibit: CMBS and CLO Issuance $bn (from Goldman Sachs Investor Presentation)

Overview

We expect bank net interest margins to increase from the current multi-decade low 3.2% (on average across FDIC-reporting banks – see Exhibit 1) to 4% in two years. This represents a 25% increase in net interest income (plus growth in earning assets) which, in turn, represents approximately one-half of the revenues for a typical regional bank. There are no marginal costs associated with margin expansion.

Exhibit 1: Reported Net Interest Margins (as % of Average Earning Assets) for FDIC-reporting US banks

Our thesis is based on the forward markets for the yield curve (actually the swap curve since we care about bank, not Government, funding). These are indicating that, by September 2015, 6-month Libor will be at 1.6% (compared to 0.4% today) and the slope between the 3-year swap rate and 6-month Libor will be at 1.1% (compared to 0.6% today). History suggests banks capture one half of the increase in short rates in their deposit spreads (i.e. the difference, weighted for deposit balances, between 6-month Libor and the cost of deposit funding) and the full amount of curve-steepening (see Exhibit 2).

Exhibit 2: Spread Contributions to Net Interest Margin for FDIC-Reporting Banks

If this history repeats, the rate environment will lift bank normalized net interest margins by 1.1% by September 2015. There is an offset in that loan spreads, currently at 2.7%, will likely come in. However, to entirely wipe out the benefit of the rate environment, loan spreads would need to fall to the multi-decade lows of 1.6% which occurred in 2005 when competition from securitization products was ramping up. The financing markets are beginning to recover (see Exhibit 3) but we do not expect competition to approach the levels of 2005-2007 particularly given that current securitizations are more “vanilla” than some of the leveraged products of 2005-2007.

Exhibit 3: CMBS and CLO Issuance $bn (from Goldman Sachs Investor Presentation)

Even assuming loan spreads fall to a long-run average of 2.2% (low by historical standards), the lift to normalized bank net interest margins is still 60 basis points. In practice, reported margins will rise by more than this because, when short-term rates rise, there is a lagged repricing of deposits so that bank margins overshoot normalized levels.

Risks

Risks to the analysis are:

  1. Outcome interest rates differ meaningfully from those embedded in the forward markets. This is likely but we note that, by definition, you can hedge forward rates at no cost (beyond the possible need to post collateral).
  2. Our model (and particularly the decomposition into deposit spreads, loan spreads, and asset-liability management or ALM spread) is mis-specified. This is likely but the error term, labeled as “Active ALM/Other” in Exhibit 2, is not a meaningful driver of year-on-year variation in reported net interest margins (although, as you would expect, the quarterly noise is greater).
  3. The impact on bank net interest margins may not be as significant as our model suggests because of offsetting factors and/or ongoing secular decline. We note that bank margins increased from 3.4% to 3.7% from 2008 to 2011 as the yield curve steepened and have compressed since then as the yield curve flattened with quantitative easing; we believe it is reasonable to assume, net interest margins will increase again as the curve responds to the wind-down of Fed tapering.
  4. We may repeat history from 2003 to 2006 when short rates increased meaningfully yet bank net interest margins were flat to down. From 2003 to 2006, the benefit from rising short rates was offset by a flattening yield curve which had actually inverted by 2006; today’s forward markets, in contrast, are suggesting short rates will rise and the curve will steepen. This is margin nirvana.
Print Friendly, PDF & Email