Background. Frost s Moat: Very Cheap Deposit Base

Transcription

1 Background Frost is a conservatively run Texas bank that was established in 1868 in San Antonio. Today it s the 52 nd largest bank in the U.S. and the largest Texan bank. Frost has $32 billion in assets, 142 branches and 4000 employees. Frost has mainly grown organically, however, they has done some acquisitions in the past where they did a good job in terms of the bank acquired and a mediocre job in terms of the price paid. The banks historical leverage ratio is 10 times, meaning that for every $1 in equity there were $10 in assets. Because many banks have much higher leverage ratios, managements strengths and weaknesses will be magnified more. A bank that s trading at 5 times earnings, but is highly levered, might not be all that cheap because a couple mistakes could take over the bank. Frost was run by family members for a long time, but present management is mostly people whom Frost has trained over the years. Only one of the executive officers is part of the family. All of management has had a long tenure, Frost s CEO for example, Philip Green, has worked with the company for almost 40 years. He was previously CFO from 1995 to 2015 and he joined Frost in Because all of management has had a long tenure, management is also elder, with the youngest executive being over 50 years old and the average age being over 60. Frost s culture is all about relationships versus transactions and because they prefer training people in-house at what they call Frost University it s likely culture won t change much as management slowly gets replaced over the years. Frost s Moat: Very Cheap Deposit Base A bank has two kinds of costs, funding costs, which is the interest it pays on deposits from customers and operating costs. Operating costs are related to the day-to-day running of the business, salaries, marketing, rent, etc. We re going to focus on funding costs first, this is what Frost has built its moat around. Most of Frost s balance sheet is funded by core deposits, half of deposits come from consumers and the other half come from commercial customers. Core deposits are industry lingo for saving and interest checking accounts, and non-interest-bearing accounts. Many banks may fund up to half of their balance sheet through other sources like money market accounts, certificates of deposit (CDs) or brokered deposits. In all of these the depositor is shopping for rates. Core deposits are much stickier sources of funding and the least likely to dry up in periods of economic slow-down. CDs will tend to pay in comparison with savings and interest checking depositors a much higher percentage of the Fed Funds Rate, and once the money matures the depositor will be once again shopping for rates. Frost is not like most banks, two-thirds of their deposits come from only two sources: non-interest-bearing and saving & interest checking. 42% of Frost s deposits are non-interest-bearing, they re totally free. Savings & interest checking cost about 0.1 times the Fed Funds Rate. In the U.S. banking industry as a whole any given bank may have slightly less than a third of deposits in which the bank isn t paying any interest. Frost has a unique advantage on getting customers to leave deposits with them over time and not expect any interest in return. In 2007 the Fed Funds rate was 5% and Frost s savings and checking accounts cost 0.47%, so 10% of the Fed funds Rate. Most banks fund themselves through money-market accounts, CDs or time deposits which costed 3-4%. Frost s non-interest-bearing deposits have grown over time. In 1995 they accounted for 28% of deposits, in 2000 they accounted for 31% of deposits, in 2005 they accounted for 37% of deposits, in 2010 they accounted for 36% of deposits and today the account for 42% of deposits. Savings and interestchecking accounts as a percentage of deposits has also remained very stable, at 25%. Money market accounts have increased by 10% as a percentage of total deposits over the same period, meanwhile time deposits have decreased from about 15% in 1995 to about 2% today. Frost s substantial decline in time deposits is likely given in a significant part to the low interest rate environment we have experienced. Because people don t see the reason why they would lock money aside earning little interest. Nevertheless, Frost s core deposit advantage hasn t gone away.

2 The 4 biggest banks in Texas: JPMorgan Chase, Bank of America, Wells Fargo and BBVA have over half of deposits, meaning the remaining 247 institutions have less than half of Texas s deposits. 30% of Frost s deposits come from San Antonio. San Antonio is the town where Frost began and where it has 22% market share. Another 26% come from Fort Worth/Dallas, where they only have 2.5% market share. Dallas is a much bigger city. 18% come from Houston, where Frost has 2% market share. 12% come from Austin where Frost has 7% market share. Frost has grown deposits at a rate of more than 9% annually. About half of that growth comes from existing customers and the other half comes from new customers. Since 2008 Frost has on average added $1.6 billion in new deposits every year. This is a significant source of liquidity. The reason Frost pays so little in interest is because it focuses heavily on the customer service and building relationships. The way Frost looks at increasing deposits is through building relationships. What Frost entails by relationship is the following: (i) the business will have an account for the day-to-day running of the business (ii) Frost will be their credit facility (iii) they ll also have accounts of a few senior officers. Frost has also strengthened their relationships by layering additional services like treasury management and asset management to cater to their commercial customers. Most consumers and businesses don t switch primary banks unless there is a very strong reason to do so. Customers may have mortgages, loans, employer salaries set up or recurring payments of all sorts in which the bank is involved. People tend to leave deposits in a bank, and over time they ll grow allowing the bank to re-invest those deposits. According to some reports the average U.S. adult has had the same primary checking accounts for almost 16 years. More interestingly, 61% of people who switched wasn t because of dissatisfaction but because they were moving, changing jobs or had changed marital status. Fees seem to come in number two as the cause customers switch banks. Frost doesn t have high fees. Frost doubles down on building relationships and catering to their customers by offering additional services. Prosperity, for example, doesn t offer these kinds of services. Prosperity s is run much more like a regional bank and they also have a lot of accounts with a small value. Their average loan in 2014 was $157,000. Frost has about $2.6 billion (20%) in loans which are over $10 million. Frost has long targeted small to medium sized business, between $10 to $100 million in sales. In 2007 there were over one million small business in Texas. Frost had 13% share of small to medium sized businesses in Frost has almost tripled deposits since 2006 so their market share there is likely to be significantly higher. Deposits in Texas increased 50% over the same time period. If Frost s deposits grow quicker than the overall growth of deposits in Texas it means Frost is increasing market share. In 2007 Frost did completed a study in which they were looking for businesses in Texas with whom they didn t have a relationship but that would probably be attracted to their brand. They found about 25,000 of those businesses and they ve been working on the list one by one trying to build relationships. All in all, Frost has 91% retention whilst paying a very low percentage of the Fed Funds Rate because of the relationships it has developed and strengthened over the years and because of the sound and safe brand they ve built since the bank was created, 151 years ago. Frost is very much seen as a sound and safe institution in Texas. It may take longer to build relationships, but it has paid off in the long-term. Frost s historical funding cost is 0.4 times the Fed Funds Rate. So, if the Fed Funds Rate is 3%, Frost s deposits cost 1.2%. However, in 2011, as a result of the Dodd-Frank act banks are no longer prohibited of paying interest on commercial demand deposits (commercial non-interest-bearing deposits). So, it s likely Frost will have to pay something for these deposits. About 60% of non-interest-bearing deposits are from commercial customers. That s 24% of deposits. Most of Frost s savings and interest checking accounts are mostly accounts belonging to businesses, and Frost pays about 0.1 times the Fed Funds Rate on these accounts. So, we can use that as a proxy to see how much they ll have to pay on the commercial demand deposits. U.S. banks pay around 0.7 times the Fed Funds Rate on money market accounts. Commercial deposits tend to be bigger than consumer deposits, so Frost has to have competitive rates to

3 some extent. So, if we assume commercial demand deposits will cost 0.5 times the Fed Funds Rate, even though they re probably going to be lower than that, we can conservatively estimate Frost s earning power when rates are 3%. Let s assume the cost of interest-bearing deposits will be 0.7 times the Fed Funds Rate. Commercial demand deposits will cost 0.5 times the Fed Funds Rate, and the rest of deposits will be non-interest bearing as they have been. Using Frost s historical net interest spread, which has been very stable, we assume the yield on securities and loans will be 6.04%. Let s also assume 0.45% in charge-offs a year, significantly higher than the 0.30% average in the last 18 years. We use higher charge-offs than Frost history warrants to allow for a margin of safety. Texas didn t experience as severe of a recession as the U.S. did in 2008, there was no housing bubble in Texas. Because Frost has 3 different types of funding, we have to calculate the weighted average return on earning assets to consider the different returns on earning assets associated with the different costs of funding, with their respective weightings. The return on earnings assets comes to 2.63%. Frost has $29.6 billion in earnings assets, this translates into $614 million in after-tax normal earnings or $9.67 per share. Right now, Frost is earning $6.90 a share. Loan Portfolio Frost has $11,3 billion in loans. Unlike most of Frost s peers that may have 70-80% of loans in real estate, Frost s total real estate portfolio (consumer and commercial) is only 48% of the portfolio. Instead Frost makes much more loans to businesses (C&I loans). C&I loans are the most relationship-based kind of lending out there. In fact, Frost stopped making mortgage loans almost 20 years ago because it became too commoditized. Customers are redirected to a company that Frost works with, but gets no fee from, to get a mortgage. Because Frost has a lot of business loans and they focus on businesses with sales between $10 and $100 million, loans tend be of a significant size. 20% of Frost s loans are over $10 million. Frost s in-house limit for loans is $25 million, but the quality of the borrower and the quality of collateral may allow for more money to be borrowed out. Before making loans, Frost tries to get the relationship with the business as described above. Essentially, this involves getting a small to mid-sized Texas business s primary deposit account, 3-4 consumer accounts associated with senior officers and other important people in the firm. This is evidenced by the fact that C&I loans are only about 10% more than C&I deposits. Frost is not a transactional bank, they re not taking deposits from one type of customers to fund loans for another type of customer. However, when it comes to commercial real estate (CRE) this is not true, about 40% of CRE loans are backed by deposits. 51% is owner-occupied, these are customers who want to mortgage the property they conduct business in. These loans tend to be safer than non-owner-occupied real estate. So, the other 40% are transactionalbased loans where the customer looks for the best terms. Owner-occupied CRE loans are similar to C&I loans and this is where Frost focuses. This compares with about 31% of Southside s real estate loans which are owner-occupied and 46% of Prosperity s real estate loans which are owner-occupied. First Financial s real estate loans are primarily owner-occupied, according to their k. Most C&I are collateralized by accounts receivable and inventory. CRE undergo similar underwriting standards to C&I loans but are usually bigger than C&I loans and are typically secured by real estate. As with C&I loans, payment of the loan is usually dependent on the operation of the business conducted in the property. C&I are usually riskier than CRE loans, both are dependent on the success of a business, but C&I tends to be collateralized by soft collateral (ex: accounts receivable) whereas CRE loans are secured by hard collateral (ex: real estate). Loans secured by hard collateral are deemed to be safer and therefore commend a lower yield. Each borrower situation may be different, but that s the rule of thumb. Frost has long been an energy lender. 11% of their loans are made out to energy businesses. 9% in production and the rest in services, transportation and other oil-related businesses. Energy is a big part

4 of Texas s economy, about 10%. Frost s energy portfolio concentration has come down over the years, not many years ago it was 16%. Nevertheless, Frost has been a very prudent lender overall and particularly so in energy. The criteria Frost uses to make energy loans start by coming up with a price deck. For example, in 2015 Frost s price-deck projection of oil prices per barrel was $50 moving upwards through 2019 and reaching $70. The borrowing base will be 65% of the discounted cash flow stream that results from the base case price of the price deck. In 2015 the price of oil Frost used to establish loans was $52. According to Frost s 10-k: Loan commitments generally must also be 100% covered by the discounted future net revenue of the reserves when stressed at 75% of our base case price assumptions. In other words, in 2015 the net revenue of the reserves assuming a $39 (0.75 * 52) price per barrel of oil should be enough to cover the whole loan value. Furthermore, the borrower's debt to EBITDA should generally not be more than 3.5. The collateral on production loans is oil and gas interests of the borrowers. Reserves must be proven, developed and producing to qualify as collateral. For the borrowers Frost deems the safest, collateral may include a maximum of 20% proven but non-producing reserves. This is pretty conservative considering Frost is determining the borrowing base solely on the cash flows the business will generate not considering the substantial amount of liquid assets these businesses tend to have. These tend to be bigger companies, as shown by the fact that energy loans tend to be bigger than $10 million. Furthermore, a portion of these businesses have hedges in place. For example, in 2015 when WTI crude oil was around $50 a barrel, 41% of Frost s borrowers were hedged at an average price of $89.50 a barrel. In 2014 Frost did a stress test on certain customers. The assumed the price of oil would be $37 in 2015 and maintained a number under $50 through They recognized that some borrowers used hedges but ignored the different cost structures borrowers had. They did this on 90% of their outstanding production-based loans, which were a little over $1 billion. The result reveals the potential exposure of approximately 7%. However, when Frost considered other assets beyond ongoing production, potential exposure was less than 1%. So, to some extent but to a lesser degree in recent years due to a decrease in the percentage of loans to energy companies Frost is vulnerable to the prices of oil. There s a negative correlation between oil prices and charge-offs but it s a very weak negative correlation. This makes sense considering that most of Frost s loans are not energy loans and also because Texas isn t nearly as dependent on energy as it was 30 years ago. Today energy is about 10% of the Texas economy today. 0.80% 0.70% 0.60% 0.50% 0.40% 0.30% 0.20% 0.10% 0.00% Frost Charge-offs Average Closing Price of WTI Crude Oil $ $ $80.00 $60.00 $40.00 $20.00 $0.00 Frost has also got some construction loans. About 8% of loans. For construction loans Frost usually requires the developer and builder to have an existing relationship with them and a proven track record of success.

5 Two thirds of Frost s loans have a floating or adjustable rate and 76% mature within 5 years. About 60% of Southside s and Prosperity s loan are also floating or adjustable rate. Frost s average loan yield isn t great. Frost has a median loan yield of 6.2% since Prosperity s is 7%. First Financial s is 6.9%. Wells Fargo s is 7.2%. Frost isn t a great lender and it s not an aggressive lender either, their loan to deposit ratio is 50%. Frost s loan yield is about 4.36% right now. Loan yields have historically been higher than yields on securities. But Frost has about $4 billion in loans that matured in 2018 and they will be able to re-deploy them at higher rates, increasing the $534 million in interest income from loans they got in So, Frost has flexibility here, but they re an interest rate sensitive bank. In 2007 the Fed Funds Rate averaged 5% and Frost s average yield on loans was 7.76%. Securities were yielding 5.24% in 2007 Loans tend to have a spread over the Fed Funds Rate. But the spread (or the premium ) over the Fed Funds Rate decreases as the Fed Funds Rate increases. Many banks may have higher yield on loans, but Frost s low cost of deposits means their net interest spread (interest income minus cost of funds) will be higher than peers. Frost s net interest spread has averaged about 3.97% over the last 25 years whereas Prosperity s has averaged 3.6% even though Prosperity s average yield on loans is 7%. Frost has $157 million in non-performing assets (NPAs), or 0.47% of assets. NPAs include non-performing loans, restructured loans and foreclosed assets. Frost has $80 million in non-performing loans. A loan becomes non-performing if the borrower has stopped making mandatory payments for 90 days. A loan is charged-off if there is no recovery potential. Net charge-offs is the amount of the loan charged-off minus the value of any repossessed assets. Frost s net charge-offs to average loans has averaged 0.27% since 2000, this compares to average industry net charge-offs of 0.91% to total loans since However, Texas banks have historically charged-off less loans than the U.S. banking industry as a whole. Net chargeoffs in Texas since 2000 were 0.37%. Prosperity, for example, has charged-off an average of 0.14% of loans. This could for various reasons: (i) because Prosperity sells a lot of the loans they get when they buy a bank (ii) most of their loans are mortgage loans, and half are owner-occupied. Mortgage loans have historically had lower charge-offs than CRE and C&I loans. Mortgage loans are backed by hard collateral (iii) Prosperity is better at making loans. When talking about the lending side of a bank, we can also look at loan loss reserves to get an idea of the quality of management. All banks have an account called loan loss reserves, which acts as a cushion for loans that are charged-off. Charge-offs will reduce the loan loss reserve and loan loss provisions will increase it. The loan loss reserve is determined at the beginning of the year and can change during the year. Frost s loan loss reserve in 2017 was $155 million. Earnings will be charged by the amount of the reserve and some banks will use this accounting treatment to magnify earnings in bad years. Frost s loan loss reserve has always been conservative. For example, in 2017 Frost s charged-off 0.27% of loans, and their allowance for loan losses to total loans was 1.18%. Frost s allowance for has had a comfortable difference between their net charge-offs and allowance for loan losses. Frost is an above average bank in terms of charging-off loans. In 2008 Frost s net charge-offs were 0.58%. Frost would need to charge-off 20% of their loans and not get anything back from collateral to wipe out their tangible common equity (shareholder equity minus intangibles). In the last 20 years the highest number of net charge-offs Frost has had to deal with was 0.67% in Frost may have a lower gross loan yield, but it also has lower charge-offs than peers. Frost s loan portfolio is 40% of earning assets. Frost s loan to deposit ratio is 52%, which is lower than Frost targets. For every dollar in deposits fifty-two cents are being loaned. Frost wants to let that trend higher, but Frost needs to find the right loans, and it may not be easy to raise that ratio if deposits grow a lot.

6 Securities Portfolio Frost has $11,9 billion in securities. Frost would prefer to make more loans and buy less bonds, but they can only make so many loans that fit into their risk profile and because of that they ve been putting the money to work in securities. Securities make up 40% of earning assets versus 38% for loans. 65% of securities are municipal bonds, 28% are U.S. Treasury and 6% are mortgage-backed securities. About 98% of the municipal bonds were issued by the State of Texas and about 70% of those are guaranteed by the Texas Permanent School Fund, which has a triple-a rating, or secured by U.S. Treasury. Frost likes municipals better than treasury and MBS because the Fed doesn t intervene in the municipal bond market in the way it does for the latter asset classes. Frost tends to buy municipal bonds with high coupons and 10-year calls. So, Frost s duration in terms of municipals to call is 4.8. If none of the municipal bonds aren t called duration would probably be around 7 years or so. Frost has also bought some shorterterm U.S. treasuries to off-set the duration. All of Frost s mortgage-backed securities (MBS) were issued by U.S. government agencies and corporations, and they never bought subprime MBS. Much like the loan portfolio, 37% of Frost s securities mature in 5 years. As with loans Frost will be able to redeploy this capital into bonds as yields go up. The portfolio duration is 4.8, which is higher than the usual duration of 3 years Frost has kept it at. The duration measures two things, (i) the time to recoup the cost of the portfolio and (ii) the change in value resulting from a 1% move in interest rates. So, if interest rates went up 1% the bond portfolio would decrease by 5% in value. The duration is higher than it s been historically because Frost is a very prudent lender and because they can t lend the money as fast as they group deposits, they ve bought municipal securities which has made them more vulnerable to the inevitable bond re-pricing as interest rates move upwards. Under the regulation known as Basel III, banks with less than $250 billion in assets or less than $10 billion in foreign exposure are allowed to be exempt from unrealized gains and losses triggering capitalization problems. So, while Frost s bonds may re-price to allow for higher yields, and therefore Frost s securities portfolio fair value will decrease, but it won t trigger a need to raise additional capital. Nevertheless, this is just an accounting issue. The effect of the paper-losses as bonds re-price will be mitigated by increasing yields on both loans and securities as time goes by. Frost s Costs We have discussed one kind of cost, the funding cost. Now we ll discuss the operating costs. Operating cost is noninterest expense minus noninterest income over earning assets. Deposits per branch are an important determinant of overall costs because high deposits per branch create a lot of economies of scale at the branch level. Frost has $200 million in deposits per branch, more than probably any bank if you exclude non-consumer and retail banks that have a big deposit base and very few branches. Wells Fargo has about $160 million, Prosperity has $74 million. High deposits per branch are an important metric when considering operating costs. High deposits per branch creates a lot of economies of scale. To put it in context, Frost can spend much more on advertising whilst hurting margins much less than competitors. If the Fed Funds Rate was 3%, Frost would be doing about $6 million in pre-tax profit per branch. If Frost spends $200,000 a year on advertising per branch that s about 3% of their pre-tax profit whilst Prosperity, for example, who would be doing about $2 million per branch would be set back by 10%. There s a lot of economies of scale from advertising, to rent, to salary expenses. In 2017 Frost s operating costs were 1.4%. In 2005 they were 2.2%. This compares with operating costs of roughly 1.9% for Dallas banks. Noninterest expense for the average Dallas bank is 3.4% versus Frost s

7 2.6%. However, noninterest income at the average Dallas bank is 1.5% versus Frost s 1.1%. Frost s doesn t have high fees and it also doesn t cross sell a lot like Wells Fargo. International Bancshares and First Financial, a similar bank to Frost, have higher noninterest income even though they re not bigger banks. Bigger banks tend to have higher noninterest income than smaller banks, according to FDIC data, institutions with over $10 billion in assets have 1.7% in noninterest income to assets. Unlike Frost some other banks have increased costs as they ve grown. Frost has consistently lowered operating costs, due in part to high deposits per branch and also due to the fact that the bank has become bigger overtime. In 2005 operating costs were 2.32%, and operating costs haven t declined because noninterest income went up. Noninterest income has actually decreased over time, partially offset by a consistent decrease in noninterest expense. The insurance and wealth management business Frost set up is in part to try to grow noninterest income overtime. Conclusion If Frost s advantage had to be nailed down to one thing, it would be the sticky and low-cost deposit base that will allow Frost to earn above average returns on assets. Across the board banks are not being priced by the market in terms of what the underlying fundamentals would render. Banks are very sensitive to interest rates. In a low interest rate environment banks are probably going to earn subpar returns on both securities and loans, yet operating costs won t decrease with the Fed Funds Rate. So, their margins are compressed to a point that even though a bank like Frost has two and a half times more deposits than in 2008, earnings have only grown 30%. A bank like Prosperity that does a lot of mortgage lending isn t as sensitive to interest rates as Frost is. Frost makes a lot of short-term loans to businesses and their gross loan yield isn t above average, their noninterest income isn t high either and Frost has a lot of non-interest-bearing deposits. For Frost to earn high returns on equity the Fed Funds Rate needs to be 3%. Frost is a very good deposit gatherer and it is in a very good state, it s grown deposits per branch at 4.3% annually since Frost spends about $1 million to get a branch up and going. Within about 27 months Frost has $35 million in deposits and if they loan out a third of that they re breakeven. Prosperity isn t as good of a deposit gatherer and they don t have such a strong brand. Prosperity takes about 3-5 years to get $20 million in deposits. Frost has grown deposits at 9.5% annually compared with 6.2% for all FDIC- Insured institutions since There s a reason Frost gathers better deposit than peers. In the last few years Frost has increased advertising spending a lot, but Frost was already a good deposit gatherer before that. They have excellent managers and relationship officers that try to increase customer retention through a variety of ways. For example, on average banks only retain 70% of customers during the first year they joined, so Frost has an onboarding program and switch kit specialists where they contact these first-year customers to help them adjust to the new bank, move automatic transactions like automatic payments or direct deposits. Other things include 24/7 customer service answered by someone from Texas, doesn t put hold on checks, over 1300 ATMs in Texas. Beyond that there is also the asset management (trust, 401k), treasury management and insurance services that solidify the relationships. A lot of banks depend on CDs and brokered deposits to fund their balance sheets, Frost s is primarily funded by non-interest-bearing deposits and savings and interest checking accounts. Together they cost very little an allow Frost to have an average net interest spread of 3.97% (average yield on earning assets minus cost funding), even though Frost doesn t have high gross loan yields. CDs and brokered deposits are looking for yield. When the deposit expires, they ll go shop somewhere else for the highest yield. It s not nearly as sustainable or cheap as Frost s, banks will pay 80-90% of the Fed Funds Rate for these deposits, versus 10% Fed Funds Rate for Frost s savings and interest checking.

8 Banks take in deposits and make loans at market rates. The only thing they can control is the quality of loans and the amount of expenses. Frost has managed to consistently decrease net noninterest expenses (noninterest expenses minus noninterest income), and it wasn t through the increase of noninterest income. As they ve become a bigger company and because they have a lot of deposits per branch costs have come down as a percentage of earning assets over the years. Frost s operating cost is 1.4%. On 42% of deposits ((non-interest-bearing) Frost is essentially borrowing money at 1.4% and putting half of it into securities and the other half into loans. The thing that can affect Frost s durability is the quality of the loans it makes. Frost has a very conservative lending culture. Someone who s looking at Frost s portfolio might worry at the energy lending and the amount of loans made to businesses that aren t backed by hard-collateral. Where Frost doesn t have hard collateral, it has a relationship with the borrower. These borrowers are involved with Frost in many other ways, they have accounts there, recurring payments, personal loans, savings that are invested amongst a number of other things. Frost isn t the place where one-time borrowers go to finance their businesses. Frost s loan portfolio doesn t have the kind of risk that a transactional bank with the same loans would have. They also don t do mortgage or credit card lending. Frost s gross loan yield isn t high, they re not aggressive lenders. Looking at managements history, it s fair to say that it s more likely the lending culture will remain the same rather than change, in other words, Frost yield on loans won t be high compared to peers but their charge-offs will be below average. Frost also has a lot of liquidity. They have more than $3.5 billion at the Fed. In 2017 there was another $3.8 billion in loans that were maturing, $534 million in interest income from securities and $315 million in interest income from securities. Frost never needed help from the government at any point in their history. The 1980s was probably the worse period banks in Texas have experienced, Texas was very oil dependent and oil prices collapsed from $35 at its peak to $10 in Frost made a lot of energy loans, so it was affected by the price drop very quickly. In 1988, 100 of Texas s 1400 banks went bankrupt. Frost is the only bank out of the 10 biggest Texas based banks at the time that existed as an independent entity during that period and until today. The other 9 banks either merged with banks out-of-state or failed. Fed Funds Rate and Deposit Growth (5-year average) 14.0% 12.0% 10.0% 8.0% 6.0% 4.0% 2.0% 0.0% Deposit Growth Fed Funds Rate Deposits tend to grow at a similar rate to nominal GDP. Since 1998 Texas s GDP has grown 1% faster than the U.S. generally, so we can expect at least 5% in nominal GDP growth. Texas is the 2 nd largest economy in the U.S. and the 10 th largest in the world. Texas has a lower cost of living than other places in the U.S. and it has a good climate, this will attract more people and immigrants to the state. Population growth in Texas has been quicker than in the U.S. generally. In the last 20 years it s been almost 1% faster than U.S. population growth. These are tailwinds that are likely to persist into the future.

9 GDP and Deposit Growth (5-year Average) 14.0% 12.0% 10.0% 8.0% 6.0% 4.0% 2.0% 0.0% Deposit Growth GDP Growth Frost s long-term deposit growth, from 1993 until now, was 9.3% CAGR. If we exclude acquisitions deposit growth was 7.3% CAGR. In the long-term we can expect 7-9% deposit growth. For the investor looking to purchase the common stock it s more sensical to look at deposits per share or earning assets per share given that banks may issue shares to fund acquisitions. Deposits per share have grown at 8.8% since 1995, whilst paying out 50% of earnings. When interest rates go up, deposits don t increase as fast. In the period between the Fed Funds Rate went from 1.67% to 5.02% deposits per share grew at 4.5% CAGR. Between the Fed Funds Rate went from 0.09% to 1.83% and deposits per share grew 4.8%. So, if the Fed Funds Rate does continue to go up, we can expect 5% deposit per share growth or so. When rates are going up deposits tend to grow slower as evidenced by smoothed out 5-year graph. In 5 years, we can expect deposit growth of about 5% a year. In 2017 Frost had $29.6 billion in earning assets. Their return on earning assets considering normal interest rates of 3% is 2.64%. This results in $781 million in earnings before taxes or $12.39 a share. And $9.79 in after-tax earnings. Assuming we return to a normal interest rate environment, Frost is worth $146 with a 15x multiple and $195 with a 20x earning multiple, today. Most of Frost s assets and liabilities will re-price in less than 5 years, because most loans will mature within 5 years and the securities portfolio has a duration below 5 years. We can calculate the return we can expect in the stock with two things: (i) what do we expect the bank be worth in 5 years and (2) what the market implies the bank is worth now. In 5 years, after-tax earnings would be $787 million or $12.49 a share (todays normal earnings * 1.05^5). At a 15 P/E multiple Frost would be worth $187 a share. From today s price of $100 a share, we can expect a 17.4% return in the common stock plus a 2% dividend yield. The S&P s average historical P/E ratio is 15. From the S&P500 has returned 11%, but we won t see the same returns in a higher interest rate environment. The value of the cash-flows produced by these businesses against an increasing interest rate will decline. Therefore, businesses lose value when interest rates go up. S&P500 will probably return much less than 11% in the next 5 years. Between 1955 and 1973 the Fed Funds Rate went from 1.79% to 8.73%, the average return for the S&P500 between was 9.8% whereas between the return was 3.5%. When I bought Frost, the common stock was trading between $97 and $99. That s 15.5 times 2018 aftertax earnings. Between 1993 and 2008 the bank earned an average of 1.9% on assets and 2.4% on earning assets (2.7% in 2007). That return is significantly higher today because operating costs have decreased over time. Nevertheless, if we assume 1.9% on assets is the normal return going forward then I paid less than 10 times normal pre-tax earnings for Frost or, assuming 2.4% return on earning assets, less than 9 times pre-tax earnings.

10 In Nathan Tobik s book he said that: If a bank pays the average rate on deposits, earns the average rate on loans, avoids bad lending as has average expenses, they aren t an average bank, but an above-average bank. Well, Frost pays below-average rates on deposits, earns the average rate on loans, avoids bad lending, has below average expenses, so it s surely an above-average bank.