We are testing a detailed look at sectors within Wall Street. Today the focus is on FIG. If you have no interest in the working of FIG or investing in FIG, we suggest you run for the door. However… If this is of interest to you, take a seat, read and take notes!
To build on what we learned from modeling Apple, we’re doing a test run on detailed articles by sector to introduce and explain certain industry silos in more detail.
First up is the Financial Institutions Group, or “FIG”, which includes: Banks, Insurance Companies, Asset Managers, Diversified Financial Companies (Credit card companies and the like), Intermediaries / Securities Firms / Other Financial Companies (Custodial Banks, Exchanges, Brokers, Financial Technology, etc.)
This post will lay out a framework for understand FIG companies, starting with banks. It will do so by exploring the business model and typical operating and valuation metrics important for each major FIG subsector, in particular how they are different from other companies.
Like Energy and Real Estate, FIG is a slightly different beast from your typical EBITDA / cash flow driven companies (“Widget” companies) because the balance sheet drives the income statement, and not the other way around. The assets of FIG companies (loans, investments, cash and securities, etc.) are what generates revenue for them, in the form of interest and investment income.
Basically, FIG companies “borrow money” (i.e., source capital) cheaply and then “lend money” (invest that capital) expensively. Much (though not all) of their income is generated by the spread between those two rates of return.
Additionally, because of the sources of some of their capital (largely individual consumers), there is strict regulation surrounding what kind of assets FIG companies can and cannot hold on their balance sheet, and in what quantities.
There are a multitude of additional complexities involved with FIG companies which will not be covered here, but this should help to provide a basic understanding of how FIG companies operate and function and how people think about them.
Lets go ahead and start with banks
How Do Banks Make Money?
Net Interest Income (50-75% of revenues). The interest they receive on their interest-earning assets (loans), less the cost of: The interest they pay on their interest-bearing liabilities (deposits) The cost of bad loans (mortgages they foreclose on)
Banks hold deposits, for which they pay little interest, and use them to make loans, for which they earn as much interest as they can. They also have to eat the cost of loans that default, how much of which is dependent on the quality of said loans and what collateral (if any) is associated with them.
Besides just deposits, banks can also fund their lending activities using wholesale funding from other financial institutions, the government (i.e., the Fed), and the capital markets. Together, these comprise a bank’s interest-bearing liabilities. Equity contributions are also a source of capital, though it is usually fairly limited because of how banks deliver value to shareholders and the capital requirements associated with how they maintain their balance sheets.
The investments banks “interest-earning assets” are primarily composed of loans (mortgages, commercial financings, construction loans, etc.), but can also include investments in other sources (securities, proprietary PE-type investments, stocks and bonds, etc.) depending on a bank’s capital position.
The reason that interest expense is included “above” the top line (in net revenues) is because interest expense for a bank is analogous to COGS for a widget company. The uses of the liabilities drive bank interest expenses are fungible between being operational and being a traditional source of financing since again, their assets are their capital.
Non-interest income (25-50% of revenues):
Mostly composed of fees, but can include other fun stuff as well.
Banks charge fees for pretty much anything they can get away with – likely the best know examples are investment banks charging advisory fees and commercial banks charging lending and deposit fees (think ATM fees and the like).
Besides fees, other common sources of non-interest income include:
- Principal Transactions– for certain sources of their capital, banks are not limited solely to fixed income able to make principal investments which are slightly more risky – merchant banking, for example.
- Asset Management – Detailed further below, but also a fee – on the assets being managed
- Credit cards – Besides the loans associated with the card, banks also use them to generate interchange fees (essentially a transaction fee – the reason so many delis in the city don’t take AmEx is because their interchange fees are much higher than other issuers)
- Some investment income not included in interest income (typically from principal transactions)
- Anything else that doesn’t involve interest income.
Income Statement and Profitability Ratios
Income Statement: Now, taking the above and walking through the rest of the Income Statement, we have the following (explanations below):
Less: Interest Expense
= Net Interest Income
Add: Non-interest Income
= Total Revenues
Less: Non-interest Expense
= Pre-Tax, Pre-Provision Earnings
Less: Credit Loss Provisions
= Net Income
Non-interest Expense – Essentially SG&A. Includes compensation expense, technology and equipment, marketing and sales, etc.
Credit Loss Provisions – Banks have to assume that some portion of their loans are going to default. In anticipation of that, they set aside a certain amount of capital each period to match what they estimate the losses will be for the loans they originated.
This is charged against the I/S in the period during which the loan is originated – i.e., not when the loans actually default (if they default).
The capital they set aside and charge against their revenues goes to a reserve fund (a contra asset) on the B/S called Loan Loss Reserves – discussed below.
Key Profitability Ratios:
Net Interest Margin (NIM) – Net Interest Income / Average Earning Assets
- Higher is better
- How effectively bank is using assets to generate income
Efficiency Ratio = Non-Interest Expense / Net Revenues
- Lower is better
- Measure of operational efficiency
Return on Average Assets (ROAA) = Net Income / Avg. Assets
- Higher is better
- How effectively bank is using assets to generate income
Return on Average Common Equity (ROAE) = Net Income / Avg. Common Equity
- Higher is better
- Ability to generate returns to investors in its common stock
Loan Loss Reserves and Asset Quality
Loan Loss Reserves
As mentioned above, banks set aside a provision each period based on the loans they originated in that period. This goes to a contra asset called Loan Loss Reserves, which is basically an “emergency fund” for when loans go bad and the bank has to cover the cost of default.
Where it gets tricky is that a loan can be behind on payments and not be considered a default. The process of disposing of bad loans therefore involves the following steps:
- Loan is made
- Borrower stops repaying loan
- Up until 90 days past due, the bank accrues the interest on the loan as if it will eventually be paid back.
- After 90 days past due, the loan goes to nonaccrual (they stop assuming they will get paid back any interest) and is considered a Non-performing Loan (NPL)
- The NPL is appraised (based on the value of the collateral and any money the borrower can repay) and the expected loss from the loan is charged against the reserves (the Net Charge-offs, or NCO)
- After the loan is foreclosed, any collateral is sold, and any recaptured principal is added back to the reserve as Recoveries. In the case of mortgages or real-estate loans, the collateral is called Other Real Estate Owned (OREO) – basically all the houses the bank has repossessed that they haven’t been able to sell yet.
Each period, the reserve calculation is as follows:
Loan Loss Reserve, BOP
+ Credit Loss Provision Expense (from I/S)
= Loan Loss Reserves, EOP
Asset Quality Ratios
A banks asset quality is determined by what portion of their earning assets are not performing – i.e., not paying up. A bank with a lower portion of NPLs and Non-performing Assets (NPAs) (any earning asset that isn’t earning, whether due to default or non-payment) is going to perform better, so banks want to minimize these kinds of loans while also making sure they have enough reserves to cover the loss potentially associate with them.
NPLs / Loans and NPAs / (Loans + OREO)
- Lower is better
- Reflect the portion of assets not earning money
NCOs / Avg. Loans
- Lower is Better
- Amount of loan losses caused by customers default and lack of collateral
Loan Loss Reserves / Total Loans
- Higher is better, but too high means a bank could have money used for reserves that could be put to better use
- Indicates adequacy of size of reserves
Capital Adequacy and Regulation
To reiterate, a bank wants as many interest earning assets as possible… and it wants to earn as much interest on those assets as possible.
The problem is that, generally, loans with higher interest rates are also loans with higher risk profiles. Since the deposit base utilized by banks is one of the foundations of the financial system, there are all sorts of capital requirements regarding what a bank can and cannot use different types of capital for. Taken to an extreme, if a bank took everyone’s deposits and lost them all betting on red, then a lot of people would be SOL (the FDIC only covers up to $250k). Regulators don’t want that to happen, so they put in rules saying “You can’t bet people’s money on red, because that’s too risky, but you can invest this money in treasuries or something we think is safe”.
To determine if a bank has enough capital to handle a “worst-case scenario” (think stress tests), banks use a variety of capital ratios to determine how solvent they really are and how big the risk that they lose everybody’s money is. The numerator of these ratios is some definition of what a bank’s “safe” capital (sources of funding) is, which is divided into tiers of decreasing safety thusly:
- Tangible Common Equity (TCE) – Equity less goodwill and intangibles
- Preferred Stock
- Trust Preferred Securities (TRUPS) – A hybrid security primarily used by banks
- Loan Loss Reserves
- Subordinated Debt
While the denominator is some representation of a banks total assets (i.e., loans and other investments):
Tangible Assets (TA) – Total assets less goodwill and intangibles (i.e., assets that could be reasonably recovered in bankruptcy)
Average Tangible Assets (ATA) – Average TA over a given period
Risk-Weighted Assets (RWA) – Total assets, where each asset class is weighted based on it’s level of risk. The risk-weightings for cash, for example, is 0%, since cash is “risk free”. Most government securities are weighted at 20% ,since they are “kind of safe”, while commercial loans and ABS / MBS are weighted 100% of their value, and can be rated higher than 100% if they are below investment grade (BB).
Capital Adequacy Ratios
Tangible Capital – Simple, conservative approach to evaluating bank solvency
Tangible Equity Ratio = (a+b) / TA
Tangible Common Equity Ratio (TCE) = a / TA, essentially tells you the amount of losses a bank can take before shareholders equity goes to zero
Regulatory Ratios – Used by bank regulators to capture the difference in risk between asset classes
Tier 1 Common Capital = a / RWA
Tier 1 Risk-Based Capital = (a + b + c) / RWA
Tier 1 Leverage = (a + b + c) / ATA
Total Risk-Based Capital = (a + b + c + d + e) / RWA
As opposed to industrial companies and due to the nature of their business, banks are valued based on cash flows
to shareholders only (in contrast to cash flows to shareholders and debt holders), as debt funding is directly correlated to the bank’s assets and its profitability.
Banks tend to trade primarily based on Tangible Book Value (book value that would be available to shareholders in bankruptcy) and Earnings (P/TBV and P/E).
- P/E – Because of the capital adequacy concerns mentioned above, banks are only able to dividend a limited amount of their earnings each period to equity holders, since they may have to retain some portion of their earnings in order to improve and/or maintain their capital position. Higher multiples are driven by higher quality (i.e., consistent) earnings
- P/TBV – A representation of how many income producing assets a bank has. Higher multiples are driven by higher ROTCE ratios.
- DCFs – While DCFs are rarely used to value banks, they can be applied to an estimation of future dividends (assuming a constant capital ratio) though this method is heavily dependent on cash flow and growth assumptions.
2) Insurance Companies
How Insurers Make Money?
This can be broken down into two ways again:
1) Underwriting Income
The premium payments they receive, less:
The claims payouts they make
The operational costs associated with generating the policies that pay those claims
2) Investment Income
Money they make by investing the cash they receive from premiums before they have to pay out claims
Most of their income typically comes from investments. Insurers can and do make money from their insurance operations, but they usually price their products competitively so that they receive as many premiums as possible. Sometimes this means that they break even (or even come out negative) on a given insurance product because if they price it any more expensively then a competitor will capture that premium.
Premiums are analogous to a bank’s deposit base – they represent cash with almost no cost-of-capital (or even negative cost of capital if an insurer is able to turn an operating profit) that insurers can invest for themselves (hence why Warren Buffet loves insurance so much). They want the investment income they can make on the cash they have lying around while it’s waiting to be paid out for policies.
Life vs. P&C
The insurance industry is broadly divided into two categories: Life Insurance, and Property & Casualty (P&C) Insurance (i.e. car / house / medical insurance – anything that isn’t life insurance).
The reason for the distinction is because of the nature of the payout periods for life insurance vs. other kinds – life policies, by definition, last longer than any other type of insurance a consumer may purchase. Therefore, once they sell a policy, they know with reasonable certainty that they have the capital they get from those premiums for a fairly long amount of time, allowing them to make conservative, long-term investments that will generate more investment income than ones with lower time horizons.
P&C insurers, on the other hand, have much shorter policies and payout periods. As such, in addition to investment income, they tend to rely slightly more on the income they can generate from their actual underwriting operations because they’re churning through policies.
The total value of a given insurance policy is not static. The revenue and expenses associated with a policy (both historical and projected) can change over multiple periods. As such, there are a lot of accounting quirks associated with how insurers report their financials, mostly to do with how to reconcile the timing mismatch and estimates informing their underwriting profit. There are a boatload of variables that can affect how a policy is valued.
Example to demonstrate:
Let’s say you have a 3 year renters insurance policy, which you paid entirely up front. The insurer now has your cash in its hand, which it can use to invest, but it doesn’t actually “earn” the money for years 2 and 3 until years 2 and 3 happen, so what’s the fairest way to recognize it? And what about the associated investment income?
Then, let’s say halfway through year 2, you get robbed and they have to pay out the full value of your claim. The recognized claim expenses associated with the policy to that point had actually been nil, but the reported expenses had been estimated (based on actuarial statistics). Now the insurer has to take this actual expense amount (the claim payout) and spread it out over the 3 years, including retroactively updating the recognized year 1 expense amount.
Then on top of that it turns out the salesman who sold the claim had a clause in his contract that he would lose his unvested bonus if a certain amount of his policy sales resulted in claims, so now the commission expense associated with the policy also has to retroactively change for year 1 and the estimate for commission expense may have to be lowered for year 3.
You also have the investment income / losses, which include both realized and unrealized interest income, dividends, capital gains and losses, etc, and all the fun accounting rules that get associated with them.
Add to all this the fact that most insurers also take out their own insurance policies (called re-insurance) in order to hedge / manage their overall risk, so if your policy was reinsured it was likely ceded, or “given” to another insurer, who is actually the one now responsible for paying you (though indirectly).
Oh, and don’t forget the taxes associated with all of the above.
As a result of these complexities, a lot of understanding insurers comes down to understanding the accounting rules associated with them.
To oversimplify, there are basically three kinds of accounting insurers use:
- GAAP / IFRS – Traditional accounting required by the SEC. This focuses on trying to what an insurer “earned” in a given period so that shareholders can see that the business is healthy
- Statutory – Accounting required by state insurance regulators (insurers are primarily regulated at the state level). This focuses on the cash insurers actually receive from premiums and pay out as claims so that regulators know that an insurer will have enough cash to cover their required payouts in the future. It also informs the rules surrounding when insurers are allowed to issue dividends to their shareholders (similar to bank capital regulation)Associated with Statutory accounting is what is called Statutory Capital & Surplus (C&S) – similar to shareholder’s equity, but with some adjustments. C&S is used by regulators to determine the maximum amount of dividends that an insurer can pay out to shareholders. The differences are basically that the income or earnings added to C&S each period are closer to actual cash earnings than in GAAP. Examples of specific differences include:
- Bonds are generally recorded at amortized cost (vs. as securities)
- Acquisition costs (cost of new and renewal policies) are charged as incurred and not “as earned”
- Realized capital gains/losses resulting from changes in interest rates are deferred and amortized over the life of the associated security
- Embedded Value (EV) – While not required disclosure, EV accounting is used by life insurers as a way of determining the intrinsic value of all the policies on their books. If XYZ insurance company suddenly decides to stop doing business , then their existing policies they have would still generate premium revenue and have claims to pay for many years into the future. EV tries to estimate what the implied value of these policies would be.
Income Statement and Profitability Ratios
Basic Income Statement
Add: Assumed Premiums
= Gross Premiums
Less: Ceded Premiums
= Net Premiums
Insurers will report both Gross and Earned premiums for each of the above, the difference being that Gross Premiums represent the premiums expected to be received over the full life of a given policy, while Earned Premiums represent, predictably, the value of the premiums from a policy that an insurer actually earned over a given period based on the contracted length of the policy.
Net Premiums Earned
Add: Interest & Investment Income
= Total Revenue
Less: Losses and Loss Adjustment Expenses (LAE) Incurred
Less: Underwriting Expenses
Less: Other SG&A Expenses
= Operating Income
= Pretax Income
= Net Income
For statutory accounting purposes, net income is slightly different:
Less: Increases (Decreases) in Deferred Acquisition Costs (DAC)
= Statutory Pretax Income
= Statutory Net Income
Direct Premiums = Policies the insurance company wrote themselves
Assumed Premiums = Blocks of policies the insurer took from another insurance company (in order to provide reinsurance)
Ceded Premiums = Blocks of policies the insurer gave to another insurance company (in order to get them re-insured)
Interest & Investment Income – Similar to interest income for banks, though there is no associated interest expense in the top line
Losses and LAE = The claims the insurer actually had to pay out in a period, along with any associated adjustments to previously paid out claims
Commissions = Commissions paid for the policies generated in the period
Underwriting Expenses = Expenses associated with actually implementing the policies they have – office
Deferred Acquisition Costs (DAC) = DAC is an asset on the balance sheet that represents the expenses associated with acquiring (generating or purchasing) new policies that have been paid but not yet been incurred (since per GAAP rules they must be spread over the life of the policy). Change in DAC represents a non-cash item on the income statement, so Statutory Net Income adjusts for it in order to get a picture of what actual cash earnings are
Unearned Premium Reserve (B/S Item): Similar to a bank’s loan loss reserves (though not a contra asset), insurers create a reserve for premiums insurers they up front on multi-year policies. They create a liability called an that increases when they receive upfront premium payments and decreases when over time as they actually earn the said premiums.
Retention Ratio = NWP / GWP
- Tells how much reinsurance an insurer relies on to balance their risks
Weighted Investment Returns = Interest and Investment Income / Total Value of All Cash and Investments
- Higher is better
- Tells if an insurer is putting its money to good use
Loss & LAE Ratio = Losses & LAE Expense / NEP
- Lower is better
- Ranges between 50 and 75% for P&C
Expense Ratio = Total Expenses (Commissions + Underwriting Expense) / NEP
- Lower is better
- Typically around 25%
Combined Ratio = Loss Ratio + Expense Ratio
- Lower is better
- Typically 90-110%. Under 90 is unusual
- Underwriting margin is 1 – Combined Ratio
Reserves Ratio = (GWP or NWP) / Reserves
- Lower is safer
- Typically around 150%
Solvency Ratio = C&S / NWP
- Higher is safer
- Usually around 70%, minimum of 10-20% depending on regulator
Risk Based Capital (RBC) Ratio = Total Adjusted Capital (TAC) / RBC
- NAIC regulatory ratio, similar to bank solvency ratios
- TAC = Statutory surplus
- RBC is calculated in a similar way to RWA for banks, with different weightings given for their various investments and other assets
- Above 200% is good, anything below 150% is bad. Under 70% requires state regulators to take control unless is corrected in 90 days.
Valuing insurers is slightly different for Life vs. P&C insurers. Both generate investing income, so valuation based on balance sheet (including ROE and ROA) is important in the same way it is for banks. Because P&C insurers generate more of their income from underwriting, their valuation is also informed more by their operational performance.
- P/BV or P/ TBV – Higher for higher ROE
- P / GWP or P/ NWP – Higher for higher premiums growth
- P/E – Higher for higher quality earnings
- P/Embedded Value – Higher for higher ROEV
DCFs are also possible for insurers in a similar way to banks – they can be valued based on their expected future dividends assuming constant capital requirements.
Other FIG Sub Sectors (Yes we likely missed a few here)
Diversified or Specialty Finance
These include FinCos (payday lenders, etc.), “Non-Bank” Credit Card Companies (Visa / AmEx / MC), Mortgage REITS, Business Development Companies (BDCs), and agency (GSEs)
The main source of revenue for both banks and specialty finance companies is net interest spread earned on loans and leases, with the funding model as the primary differentiator
Whereas banks primarily extend loans by expanding credit through fractional reserve banking, i.e. “deposit funding”, specialty finance companies must secure its loanable funds in the capital markets
Deposit funding is a unique legal privilege granted to banks, but comes with significant regulatory strings attached limiting the potential scope of banks’ lending activities Hence, there is a need for specialty finance companies to deliver credit products that do not fit well within a bank regulatory construct Specialty finance companies also compete directly with banks in certain areas.
Diversified financials are valued in the same way that banks are.
Asset managers include both traditional long-only firms, like Franklin or Fidelity, and alternative asset managers like KKR or Fortress. Working on Wall Street, you should be familiar with the asset management business model. They invest (and hopefully make) money for people (or endowments, insurance companies, etc.) in return for a fee. Broadly, there are two kinds of fees asset managers can take:
- Management Fees – This is a % of the total AUM, and is paid regardless of performance
- Performance Fees – this is a % of the returns generated by the asset manager, with fees usually assessed on any performance above a given benchmark or high water mark
Because of this, one of the most important metrics for asset managers is their AUM, and how quickly it is growing (or shrinking). AUM can grow (shrink) in 3 ways:
- Investment gains or losses
- Organic Flows = Money given to or taken out of an asset manager by their clients
- Acquired Flows = AUM acquired from another asset manager
The net organic flows of an asset manager or highly predictive of it’s market value, with higher flows obviously being better.
Besides AUM and flows, however, asset managers are valued and treated similarly to other EBITDA companies, so we won’t go into more depth here.
Finally, there is “all the rest” – other financial firms that don’t fall into the categories outlined above. These are primarily in the securities industry, and are valued and analyzed in the same way as Widget companies. Examples of securities companies include:
- Brokers (online and retail)
- Boutique investment banks or advisory firms
- Execution Services Firms
- Financial Technology / Software Companies
- Market Data Providers
- Financial Processors