Description

Company Description: First Tennessee is a Tennessee based regional bank that has aggressively grown its mortgage banking and mortgage underwriting/trading operation in the last 5 years. Until 1995 First Tennessee was a decent somewhat undifferentiated regional bank with a normal bank like mix of commercial, real estate and mortgage assets. In 1995 FASB changed the rules on mortgage servicing right accounting to allow gain on sale for all mortgages originated and sold with the right to service retained by the originator. At that point FTN significantly increased its investment and dedication to the mortgage banking business (much like Countrywide). The business chugged along until 1998 when mortgage originations went from 835 billion to 1507 billion nationally (27% growth in purchase orig, 201% growth in refi originations-the 10 year declined 200bp from mid 97-end of 98). In 1998, FTN’s mortgage banking operation really picked up (both mortgage origination and mortgage backed securities underwriting and trading). More than 100% of the earnings growth in 1998 was driven by the mortgage business. FTN’s mortgage originations generally moved in line with the mortgage market shrinking 13% in 1999 (capital markets revenue shrunk 14%) consistent with the mortgage market’s 16% decline, but its revenue from hedging and servicing enabled FTN to grow revenue 8% (although earnings contribution from mortgage ops and capital markets was down approximately 24% in 1999). In 2000, FTN blew up its mortgage banking operation when its hedge failed, and it was overexpensed for the shrunken environment. Earnings went from $1.87 to $1.10. They covered this up with divestiture of random assets enabling them to print $1.76 in earnings. Here is a quote from a press release issued in Feb 28, 2000:
“Over the last several months, there have been various market factors which have adversely impacted the profitability of mortgage banking and capital markets. While the effect of these factors has been historically short term in nature, in the current economic climate they could have a more prolonged effect. The primary factor impacting mortgage banking and capital markets has been the persistent, rapid rise in interest rates and the strong universal expectation in the market that this trend will continue. As a result, mortgage banking has been unfavorably impacted by decreased loan originations and pricing pressures (especially related to wholesale originations) and other factors relating to the secondary marketing process. This expected trend in interest rates has continued to affect the level and mix of securities demanded by capital markets customers even though Y2K concerns have passed.“
The stock had already gone from $30-$35 to $23, and declined to $16.50 in the aftermath of this release. The 10 year bond went up from 6% in Sept 1999 to 6.5% in Feb 2000. To be fair, all banks sold off in this time period as the yield curve flattened with short rates catching up to long rates.

In 2001 and 2002, mortgage banking and capital markets took off again as the mortgage market exploded. Revenue from Capital Markets went from 118 in 2000 to 344 in 2001 (no acquisitions), to 447 in 2002 and likely to 563mm or so in 2003 (almost 400% increase). According to FTN’s segment reporting eps contribution from capital markets went from .20 in the 1997-2000 time period to approximately .77 in 2003 and contribution from mortgage banking went from .36 in 1998 to $2.15 in 2003 see chart below

1998 1999 2000 2001 2002 2003
Bank 1.09 1.32 1.38 1.34 1.23 .85
Mortgage Bank .36 .25 .09 .47 1.32 2.15
Cap Mkts .19 .16 .20 .62 .76 .77

% of pre corporate earnings
1998 1999 2000 2001 2002 2003
Bank 66% 76% 83% 55% 37% 23%
Mortgage 22% 14% 5% 19% 40% 57%
Cap Mkts 11.6% 9% 12% 26% 23% 20%

They are definitely sandbagging some of the numbers in the bank to offset the strong mortgage contribution and this will offset some of the decline in mortgage banking, but just using Average Returns on Assets from 1997-2002 and applying them to my guess for segment assets in 2004 gets me to an earnings number of approximately $2.35-$2.40. Being somewhat generous I can get them to the high $2.70s. FTN thinks they can do $4.00 and consensus is $3.73. I am generous in my model and say they can earn $3.08, but it is very difficult to know how quickly the capital markets business will evaporate and how quickly they can cut compensation and headcount in their capital markets and mortgage banking operations.

The Capital Markets business is approximately 2/3 fixed income trading and underwriting, bonds and MBS primarily and 1/3 trust preferred issuance, portfolio mgmt for small banks and their institutional equity business Midwest Research. Of the fixed income portion 2/3 is trading and underwriting for other banks and credit unions and 1/3 is what they call “absolute return” customers like hedge funds and insurance companies. This business has definitely gotten a significant lift from massive mortgage backed securities prepayments (which requires banks to then go buy something else) as well as limited loan demand from commercial customers forcing banks to invest deposits in securities. I believe much of it will go away, but am modeling only a 29% decline in 2004 which brings them back to a level above 2001, but below 2002. According to their segment reporting, their capital markets business earns a 5% ROA. FASB has required banks to deconsolidate Trust Preferreds from their balance sheets and the FED is currently considering whether it should maintain its existing policy of keeping trust preferreds as a component of banks core (tier 1) capital, or should align its treatment with FASB’s treatment and exclude Trust preferreds from Tier 1. If it decides to exclude Tier 1, FTN will likely lose approximately .15-.20 of annual earnings.

Thesis
First Tennessee trades pretty close to in line (1 multiple discount) with its peer companies even though in 2003, 60% of its earnings derive from the mortgage banking business which is given a 7-8x multiple at Countrywide and Washington Mutual. I believe there is a high likelihood for some sort of miss in 2004 if mortgage refinance volume dries up due to rising interest rates or even rates stabilizing at today’s rates. The company is earning ridiculously high returns in its capital markets business and its mortgage banking business which are neither normal nor sustainable. Even if the company does the $3.77 that analysts estimate-which I believe would require the company to get every hedge right and massively cut expenses the stock should trade at 12x or $45. If the company significantly misses estimates it could easily trade down to the low $30’s based on 10-11x my approximately $3.08 estimate which is typically where banks that miss trade. FTN trades at over 3x tangible book.

Bull Case
· Liquidity won’t dry up in capital markets business-it is not driven by rates
· They know how to perfectly manage the mortgage business, and can cut incentive comp fast enough to make up for declines in origination volume
· They have been spending money to reduce earnings and invest in the business which they can ramp back very quickly in the event of a slowdown (this includes advertising and charitable contributions).
· Including 2000 the bank has earned an average ROE of 19% which equates to $3.00 per share

Risks to short
· There aren’t too many bulls on this stock-sellside tries to convince itself that they can continue overearning their history, but generally they are neutral for fear of what happened in 2000
· They are well capitalized
· They should have a decent third quarter

Catalyst

1Q potential miss due to compressing spreads in mortgage banking business and slowdown in fixed income mbs origination and trading