Return to Table of Contents

Chapter 1: The Basics of Underwriting

Neither a borrower nor a lender be,
For loan oft loses both itself and friend,
And borrowing dulls the edge of husbandry.

-William Shakespeare, Hamlet

The two most important numbers in real estate underwriting are the Loan to Value Ratio (LTV) and the Debt Service Coverage Ratio (DSCR).

The LTV is computed by dividing the amount of the loan by the amount of the value of the collateral, e.g. if the collateral is worth $100,000, and the loan is for $65,000 the LTV is 65%. The LTV is almost always expressed as a percentage. Obviously, the lower the LTV is, the more secure the loan is. A general guideline is that the LTV should be lower than 75%. An LTV above a certain percentage is a loan policy exception. (This is bank talk which means "by making this loan we have violated the standards of underwriting we have set for ourselves in our written policy." Check the manual for what percentage constitutes an exception.) If the LTV is below 65% or so, then low LTV can be counted as a strength. Below 50%, and the bank should never lose money on the deal. As I said, to compute the LTV requires two ingredients. It is simplicity itself to determine the loan amount. (Well, of course there actually is some complexity here. For a new loan, you have to figure out exactly what the borrower is asking for and what you might offer. Often, the loan amount will be determined by working backwards from what LTV you are willing to have for the loan, and the bank's determination of the value of the collateral.) The more interesting question is, how do you determine the value of the collateral?

The DSCR is computed by dividing the net operating income (NOI) of the collateral, whatever that is, by the total amount the borrower has to pay each year to repay the loan (this amount is referred to as the "debt service", hence debt service coverage ratio). The DSCR is almost always expressed as a ratio, e.g. 1.27X, but occasionally as a percentage, e.g. 127%. Obviously from the bank's point of view the higher the DSCR the better. A general guideline is that the bank looks for a 1.2X coverage. Computing the DSCR is more complicated than computing the LTV. A number of variables determine the yearly debt service: the amount of the loan, the interest rate, and the amortization period. Even with that information sometimes the debt service is difficult to calculate. What interest rate do you use for a variable rate loan? Generally building in a margin for rising rates is expected. Now the question is, what is the NOI, and how do I compute that?


From above, we recall that in order to compute the LTV we require the value of the collateral. And first of all, what is collateral? Collateral is whatever of value the borrower promises to give to the bank if he* is unable to repay the loan. In the case of a real estate lender like Corus, collateral for the loans is usually a piece of commercial real estate property. Commercial real estate is any real estate that is used to generate income. Rental apartments, warehouses, offices, hotels, and even nursing homes are all examples of commercial real estate. Interestingly, farms are not considered a member of this set. One easy way to find out the value of a piece of real estate is to get an appraisal. Corus gets an appraisal for every real estate loan it makes above a certain amount and failure to do so is an exception to the loan policy. (What is the amount? Read the loan policy manual.) It is also a violation of federal banking regulations. The appraisal is prepared by a certified appraiser on the bank's approved appraiser list. (Who's on the list? Check the loan policy manual. There's a theme here, isn't there?) All appraisals must contain certain elements and conform to certain regulatory standards. These elements and standards are detailed on an appraisal compliance checklist that must be filled out for each appraisal after it is received. Failure to complete the checklist is a violation of the loan policy.

An appraisal should explain the methodologies employed by the appraiser in determining the market value of the property. (Market value is a term that should be defined in the appraisal.) Appraisals should almost always contain the following three "approaches to determining value": (1) the income capitalization approach, (2) the replacement cost approach, and (3) the market comparison approach. I describe the approaches in reverse order. In determining a property's value by the market comparison approach the appraiser attempts to find comparable properties ("comps") that have been recently sold and thereby estimate what the subject property could be sold for. To determine the value of the subject from the study of comps the appraiser must reconcile for whatever variables might make the value of the subject different from the comps. These variables include: size, location, age, condition, design, and anything else the appraiser can think of.

The replacement cost approach asks what it would cost to actually build the property from scratch. This approach presents a number of difficulties. What is the price of vacant land on which to build? This is especially difficult to answer if there is no comparable vacant land in the area. Another problem is that a building's value my be decreased substantially by wear and tear, which we call depreciation. (That's not quite accurate. Depreciation also includes a loss of value that is caused by functional obsolescence even if whatever is depreciating is perfectly maintained.) Of course, there is no way to build a new building that is worn and torn. Despite these problems, for a newly constructed building in the middle of the cornfields, this method makes a lot of sense. Why would anyone buy a building for more than he would have to pay to buy vacant land across the street and have it built for him?

The third method is the most complicated, and possibly the most meaningful. It is certainly the method the bank pays the most attention to. Why? Because the income capitalization approach is concerned with the essence of financial investment: the return. People invest in real estate for the same reason they invest in stocks, bonds, futures, or lottery tickets: to make money. The final test of your investment, in hindsight, is what return you made from it. For the past several years it has been almost impossible to invest in the stock market and not see a 20% annual return. (By the way, if you want to learn how to fail to make money in the stock market, contact former Corus credit analyst Ryan Huddlestun. He'll get you started.) Back on target again: if you invest $100 at the beginning of the year, and you make a 20% return, you'll have $120 at the end. Blows your mind, doesn't it? However, it is probably unrealistic to expect this quantity of return for the long term future.

For income producing real estate, a 10% return is perhaps more realistic. Probably that is a more realistic long-term goal for the stock market. What this means is that someone who buys a building for $1 million expects to earn $100M per year (M=1,000, so $100M=$100,000) in income from the building. An appraiser (and the bank) attempts to reverse this calculation, and determine from the income produced by a property what it is worth. In the above example, the appraiser would determine that the income was $100M and divide that by 10%, the return, and determine the building was worth $1 million.

There are a couple of fancy terms that real estate people use when referring to this calculation. Instead of talking about the income, we always speak of the NOI (which I promise to explain soon). And instead of talking about return, we always speak of the capitalization rate, or the cap rate. To determine the NOI the appraiser performs the calculation I describe below in the section about the DSCR. Once he has the NOI, the appraiser computes the value of the property by dividing the NOI by the cap rate as described in the previous paragraph.

The cap rate can be interpreted as a measure of the property's riskiness. A property that is sure to produce its expected income should have a low cap rate. Conversely, a property that is risky, should have a high cap rate. Why? Consider two properties that are both expected to produce an income of $100M per year. For whatever reason, Property A is considered a lock to produce that income, whereas Property B is a gamble. The marketplace is so convinced that Property A will produce $100M in income every year it could sell for $1.1 million. To justify that value, Property A must have a cap rate of 9.1%. (Check on your calculator: $100M/9.1%=$1.1 million.) Property B, on the other hand, has investors so nervous they will only pay $900M for it. Therefore, the cap rate applicable for Property B is 11.1%.

Real estate people love talking about cap rates. You're not prepared to go into battle until you feel comfortable answering questions like "What cap rates are downtown Chicago hotels trading at?" What this question means is what is the price of hotels in relationship to their income? Since cap rates are a measure of risk they depend on property type. Generally hotels have high cap rates, residential apartment buildings low caps and office and industrial buildings somewhere in between. (Consider the above sentence. What does it imply about the relative riskiness of those types of properties? Why should this be true?) Cap rates are also a measure of the basic real estate climate. In good times cap rates will fall across the board. In bad times they'll go through the roof. And when times are really bad, there is no such thing as cap rate, since properties aren't trading at all, i.e. no one will buy your crummy buildings at any price.

After doing all three independent analyses, the appraiser should attempt to reconcile the three values. In doing so he should explain how much weight was given to each methodology and why. The appraisal should conclude with a date and a single value for the appraisal's date.

Occasionally an appraiser will employ one or more different methodologies. For instance, for a building that is wholly leased to a single tenant for a ten year term, the appraiser might consider the present value of the income stream from that lease, discounted at whatever rate he feels appropriate, and then adding the residual value of the building at the end of the lease term. I hope that without any comment on this methodology on my part you can see from that one sentence description why that method is more problematic than the previous three: there are too many future events and conditions to guess about. (I snuck in the term "present value" above without explaining what it means. I'll cover that in about 40 pages. If you don't know just forget about it for now.)

Leaving that exhausting discussion of appraisals aside for a moment, let's think of an easier way to determine the market value of a property. An investment obviously has an associated cost: in the case of real estate investment it is generally the cost of buying or building the property. With this value you can compute a particular example of the LTV, the loan to cost ratio. The actual cost of an investment is of great importance to the bank. If an appraisal says the property is worth $6 million, and the borrower is purchasing it for $3 million, and asking for a $3 million dollar loan, then an unsuspecting credit analyst might say that the LTV was 50%-excellent! Of course this would be nonsense, the loan to cost, and hence the LTV, is actually 100%-terrible. At this point alarm bells should be ringing: how on earth can the purchase price be so far from the appraised value? Before you even think about making a loan for this transaction in any amount, you had better understand fully why there is such a great discrepancy. Another important question that should be answered is how much of her actual money is the borrower putting into the deal? This amount is called the borrower's equity. There are two big reasons for this. The first is that an equity investment keeps a borrower on the hook. After plunking down a million of his own dollars a borrower is much less likely to simply hand over a building when times are rough and walk away. The second is that a high equity investment is convincing proof that the low LTV we calculated actually has a basis in reality. And the lower the LTV, the less likely the bank is to lose money, even if the borrower does give up and walk away.


I could go on picking up loose threads from the above discussion and going in many different directions, but instead I want to shift to the other critical ratio of underwriting: the debt service coverage ratio. As I said there are two ingredients for the DSCR, the debt service and the net operating income. First the debt service. Debt service is based on the particular terms of each individual loan. There are many different variations, but there are two methods which are used most of the time. The most common is a fixed monthly payment which is sufficient to amortize the entire loan balance in a given time period. Generally in home lending the amortization period is equal to the loan term. In commercial real estate lending the loan term is shorter than the amortization period, leaving a so-called "balloon payment" of unpaid principal at the loan maturity date. The second common method is to make the borrower pay only interest during the term, and all the principal at maturity.

In order to compute the debt service, for an amortizing loan, I recommend using the PMT function on excel. Of course, if you prefer, an abacus is acceptable. The cell should have an entry that looks like:

=12*PMT(rate*(365/360)/12,amortization period*12,loan amount)

This is a discussion of underwriting and not excel, so I won't delve into the mysteries of that formula, except to comment on the (365/360) part. For almost every loan the bank makes the borrower is required to pay interest based on a 360 day year. What this means is that every day interest accrues at a rate equal to the agreed interest rate divided by 360, instead of by 365, so by the end of the 365 (or in leap year 366) day year, the borrower who thought he was paying 10.00% on a loan has actually paid interest at the rate of about 10.14%. Usury isn't just immoral, it does a little sleight of hand as well. Note: Read this paragraph as many times as it takes until you understand it. If you cannot, quit immediately.

Now comes the question I've been putting off: what is the net operating income? A property produces income through the collection of rents. It also produces various expenses: real estate taxes, maintenance, insurance, management, etc. If you subtract the total expenses a property incurs from the total rents collected, you arrive at the net operating income. Computing what this number was for a given year is easy when you are given the information from a tax return or an audited financial statement. Trying to predict what it will be in the future is more difficult, especially in the absence of historical data. Start by determining the gross potential income (GPI) of the property. The GPI is the greatest possible amount that a landlord could potentially collect from the tenants. It includes all rents that could be collected assuming the property is fully occupied at market lease rates and any expense reimbursements (whatever those are) the tenants might make. From this you should subtract a vacancy factor, generally 10%. Always make an allowance for vacancy, even if the property is fully leased. What if a tenant breaks the lease, or there is a rent collection problem, or a lease expires and the property is vacant for 5 months before it is retenanted? All these factors should be considered in making a vacancy estimate. Adding the possible rents and reimbursements and subtracting the vacancy gives you what is sometimes called the effective gross income (EGI).

Once you have the EGI you have to subtract the property's expenses. The four most basic expenses are: real estate taxes, maintenance, insurance, and management. A variety of other expenses can become relevant: advertising, decorating, scavenger (garbage pick up), utilities, exterminator, snow removal, janitor, travel, etc. For buildings that are really operating businesses like hotels and nursing homes, expenses are even more varied. (Of course, for hotels and nursing homes, the income section looks pretty different too.) Subtracting expenses from the EGI gives you the NOI. (Finally.) Please note the one thing I didn't include in the list of expenses: mortgage payments. Mortgage payments are not an operating expense of the building. Why? Because the amount of the payment is largely independent of the building: it depends on the terms of the loan, and not on the operations of the building. That is why we spend so much trouble figuring out the NOI: to figure out how much debt the building can support.

What about those reimbursements? Reimbursements, or reimbursable expenses, are expenses that the landlord passes on to the tenants. A lease can contain whatever provisions the tenant and landlord care to come up with. Some retail leases are particularly Byzantine, containing provisions whereby the landlord collects a certain percentage of the gross sales after a certain dollar figure is passed. Sometimes the landlord simply charges the tenant a fixed rent no matter what the expenses to the property are. This is probably the simplest kind of lease. However it is a great risk to the landlord. What if expenses go way up? What seemed like a profitable lease could end up costing the landlord money. It is more common for the landlord to require reimbursement of at least some expenses. Real estate taxes are generally passed on. Sometimes the lease specifies fixed percentages that must be paid by the landlord and tenant, sometimes the tenant must pay all, sometimes a certain base amount must be paid by the landlord or tenant, and any increase must be paid by the other. As I said, anything they agree to is possible. If a tenant agrees to pay all the expenses, it is called a triple net lease. (There doesn't seem to be any good explanation for this name. The best I've ever heard is that the triple refers to the three expenses charged to the borrower: community area maintenance (CAM), taxes, and insurance. However, this doesn't make sense since management isn't included. And in any case, a true triple net lease requires the tenant to pay for everything, not just those expenses that are listed in the lease.) In the case of a triple net lease, the money that the landlord actually collects that does not cover an expense, the landlord's income, is referred to as the triple net rent. This is important. In comparing properties' values people often talk about the properties' triple net rents.

There is a category of expenses that exists outside of even a triple net lease: the reserves. Reserves are expenses that do not occur regularly, but only in extraordinary years: for example, replacing the roof. (Although, I've read at least one triple net lease where the tenant was responsible for roof maintenance, but this is the exception.) Since these expenses occur only in extraordinary years, some lenders, including Corus at times, require the borrower to make an extra monthly payment in addition to their loan payment to an escrow fund for the day when the extraordinary expenses arise. Hence, the name reserves. In theory, any building owner should be setting part of his yearly profits aside for these future expenses even if the owner's lender does not require it. Reserves come in three general categories: tenant improvements, leasing commissions, and structural repairs. Tenant improvements (TIs) are costs that the landlord pays for remodelling the building to get a tenant to move in. Usually this is most significant for an office tenant. People who rent office space want partition walls, carpeting, gold plated toilet seats, and what have you. To entice tenants to sign the lease a landlord has to offer to provide these things; and of course, pay for them. For particularly spectacular improvements, which a landlord feels are unlikely to have any residual value in enticing a new tenant, like gold plated toilet seats, a landlord may require a tenant to pay for them himself, or amortize the cost of them during the term of the lease in the form of extra rent. Leasing commissions are simply fees paid by the landlord to a broker who brings him a tenant. These fees can be considerable. Structural reserves are for repairs that cannot be considered routine. Drawing the line between structural reserves and maintenance can be difficult and some leases spend a long time categorizing what is expected of the landlord and what of the tenant. Generally, maintaining the roof and the parking lot is the landlord's responsibility. If you subtract the yearly reserve requirement from the NOI you get the NOI after reserves. (As opposed to the NOI before reserves. See, this isn't rocket science.) On any given year, the property's income should be close to the pro forma NOI before reserves. Why? Because reserve expenses should occur infrequently. But for one year out of several, the income will be well below that, and also well below the NOI after reserves. But averaging the regular years with the extraordinary years should produce something close to the NOI after reserves. At least, this is the theory.

Now that you know what the NOI is, you can compute the DSCR. The DSCR is the main event of a cash flow discussion, but there are a few related terms that you should be familiar with. One is the excess cash flow. The excess cash flow is the amount you get if you subtract the amount needed for debt service from the NOI; it's the amount of money that will actually make it into the owner's pocket. Obviously, you can speak of the excess cash flow either before or after reserves. People are sometimes interested in what situation will lead to the break even cash flow (i.e. the situation when the DSCR is 1.00). Some questions: What is the maximum level of vacancy at market rents which will allow the property to support the debt? What is the minimum rent required to service the debt fully occupied? These are the worst case scenarios that the property can experience that will still allow the borrower to make payments to the bank without dipping into his own funds. Coincidentally, dipping into his own funds is exactly what I will begin discussing below.

Borrowers and Guarantors

Real estate underwriting, like Gaul, can be divided into three parts. In many reports I have read, and in most that I have written, the three most important sections are the Collateral Discussion, which should always contain the LTV, the Cash Flow Discussion, which should always contain the DSCR, and the Borrower/Guarantor Discussion.

I am going to describe the difference between a borrower and a guarantor later, when I begin to describe loan documents. For now, I use the terms more or less interchangeably, and by them mean to indicate the person doing the borrowing, the customer. (Sometimes the customer is not a person. For instance, a corporation might be the customer. This discussion will only be concerned with human customers.)

A property has two characteristics that it is critical for an underwriter to understand: its value and its income. (LTV and DSCR, see?) The same applies to the customer, but we refer to a person's value as his net worth. The basis for computing a customer's net worth is his personal financial statement (PFS). This is essentially a balance sheet which lists all the person's assets and liabilities. Surprisingly, many people exaggerate their net worth. A house purchased for $150M a year ago is now worth $250M. That real estate portfolio of two decrepit 3 flats built in 1926 is worth a cool $700M. My 25% interest in the family plumbing business: $400M. The 1992 Chevy Impala: $20M. For this reason, a credit analyst has to go through the list of assets and by making a series of adjustments to the customer's financial statement arrive at a more realistic figure of the customer's net worth.

There are at least two types of assets which are easy to value: cash and marketable securities (stocks & bonds). The only question about them is whether you believe they actually exist as the borrower describes. (One quick test is to compare the customer's tax return with his PFS. If he claims to have a stock portfolio of $1 million and shows dividend income of $436 on the tax return, alarm bells should ring. Unless he invested in internet stocks.) Real estate assets can be verified by the bank by a variety of methods. Asking for appraisals, or tax returns for the entities that own the properties, or inspection of the properties listed, or any of the things that go into real estate underwriting. Ownership interests in privately held companies are extremely difficult to value. Certainly a business that provides the customer with a stable $150M income per year must be worth something. However, the value of the company is contained in the people running it. If the customer and his partners/family quit, the business could very quickly be worth nothing, or less, if there are remaining debts. Residential property is certainly worth something, but if there is already an 80% home mortgage on the property, it is unlikely that the customer will be able to raise much money to pay his debts from this source. Personal property like cars, color televisions, stamp collections, and "gold -plated-digital-escargot forks" (Whoopi Goldberg) should be discounted at a conservative rate of, say, 100%. Not only is a borrower unlikely to be able to raise much money selling the 1992 Chevy Impala, he's unlikely to want to. More likely, you'll foreclose long before a borrower starts selling personal property.

After determining a realistic figure for the person's net worth a credit analyst's attention turns to the customer's income. You can determine someone's income by reading it off his tax return. (There is a form you can fill out and get signed by the customer to make sure that the tax return submitted by the customer is the same as the tax return submitted to the IRS. The presumption the bank makes is that nobody will ever inflate his income when filing with the IRS. However as with the above discussion, I assume we believe the customer to be honest.) It would be nice to have two or even three or even four years of tax returns, to make certain that the income level is consistent, or even better, rising. Watch out for extraordinary items, such as sales of property or inheritances, which can inflate income for one year. Ideally a real estate owner should have a steady stream of income from his developments and holdings. Be advised that simply reading the income from the tax return can give you a false sense of the real cash the person earned in a year. Real estate can involve considerable non-cash expense such as depreciation. To get a true sense of the actual cash income the analysis should include adjustments, or "add-backs", of the non-cash items.

After getting a sense of what the customer is worth, and what the customer makes, the analyst is still left with a big question: how does the customer's income and net worth contribute to the soundness of the loan. A person with a net worth of $1 million, including liquid assets of $500M, and a real property with a recent appraised value of another $500M, is a great guarantor for a $350M loan, and a lousy one for a $15 million loan. Ideally, all borrowers would have liquid resources to repay their loans immediately, hence the old jibe about bankers only lending money to people who don't need it. Of course such strong customers are extremely rare, and we sometimes believe LaSalle has hunted them to extinction. As a rule of thumb you would like the borrower to have sufficient resources to be able to carry the property for some period of time. Problem loans are in my experience often traceable to a tight but sufficient DSCR, an acceptable but aggressive LTV, and a weak borrower. Any property is going to have problems that will require infusions of capital. In time these repairs should pay for themselves, but that doesn't help a borrower who needs $25M, or perhaps $2 million, today. If a borrower is stretched thin: his only liquid capital is a checking account, or he's taking a second mortgage on his residence to make the equity contribution to the transaction you're underwriting, or his entire portfolio is highly leveraged, it's a time to worry. Another huge lending bloomer is to put a mortgage on a building that a borrower owns free & clear for his living expenses. The problem here is that if the income from the building is insufficient to support the person's lifestyle, how on earth will it produce enough income for the person to live on after first paying the bank's loan? Someone who makes that request is simply living beyond his means. That is a borrower you do business with at your peril.

Determining how much extra security a particular guarantor brings to a loan is a very difficult question. Not only because it's so much harder to determine what a person is worth than a building, but because it also requires a measurement of the person's character. Making that judgement lies beyond the scope of this discussion.

Presentations and Reviews

There are two different kinds of reports that lenders are called upon to write about loans: presentations and reviews. A presentation is for a proposed loan, and a review is a yearly report written for a loan that is already in place.

Presentations are written by officers to convince the appropriate authority to allow the loan to be made. As such a presentation is something of a sales brochure. The loan policy manual explicitly denies this is true, but I choose to disagree. A loan presentation should of course always contain a thorough discussion of any negative factors that may make a loan unsound. An officer who does not fully disclose the weaknesses of his loans to his bosses is one who deserves to be fired. But, no officer would ever go through the time and trouble of writing a loan presentation for a deal that the he did not want to see closed.

The very first page of a presentation should always contain exactly who the borrower for the loan is, exactly how much money the borrower wants, on what terms the money will be lent, and what the purpose of the loan is. Once these basics are laid out, a presentation should also contain an in depth discussion of each of the topics I discussed above. In addition, a presentation should cover the bank's previous history with the customer. A presentation should also contain a discussion of other loans the customer has at the moment. We refer to this as the bank's exposure to the customer.

Presentations are generally not long. If a presentation is more than a dozen or so pages either the deal is exceptionally complicated or the officer is exceptionally verbose. Sadly, many presentations suffer from the latter defect. However, presentations almost always contain a number of exhibits. Exhibits can sometimes quadruple the thickness of a presentation, even if the officer preparing the exhibits is not particularly prolix. (Beware of presentations and officers that fall into the underwriter's fallacy. No matter how thoroughly and well underwritten a property is, it does not increase the value of that property one cent. It seems obvious, but if you keep your ears pricked up, you'll be surprised how often you hear people fall into this trap.)

One of the most important exhibits that is almost always attached to a presentation is the commitment letter. In the commitment letter the bank sets out the exact terms and conditions under which it will make the loan. Just as it sounds, the commitment letter not only sets out the terms, but actually promises the borrower that, if the conditions it contains are met, the bank will make the loan. For this reason preparing accurate and complete commitment letters is one of the most important jobs a loan officer has. The commitment letters that Corus issues are usually extremely detailed. After the customer reads them there is often a period of negotiation about various conditions it contains beyond the basic terms of the loan. At Corus, we consider this a strength rather than a weakness. We believe it is better to argue out as much as possible before getting the lawyers involved with all their complicated documents. Sometimes, of course, lawyers do get involved with negotiating commitments, but it is usually better to avoid this. Being good at preparing and editing commitment letters is really what separates the sheep from the goats in the ranks of credit analysts.

Other exhibits that are generally attached to a presentation are: cash flows, personal financial statements, tax returns, photographs of the property, maps and site plans indicating the location and design of the property, and credit reports on the borrower. Naturally all exhibits should be referenced and discussed in the text.

Sometimes an officer will write a loan preview before writing a full-blown presentation. A loan preview is shorter and less detailed than a presentation. It gives the higher authorities an opportunity to give their input into deals before negotiations go to far, and to kill deals that the bank is not interested in before officers waste too much time on them. A preview would never have a commitment letter attached, but it might sometimes contain an application letter. An application letter is like a commitment letter inasmuch as it should contain all the terms under which the loan will be made, but it does not create a commitment on the part of the bank. Nevertheless, the bank is extremely cautious about issuing application letters. At the very least, it is poor salesmanship to issue an application letter, collect an application fee, and then refuse to make the loan. Application letters are a bit shorter than commitment letters since they do not have to contain the legal mumbo-jumbo that accompanies an actual commitment. When the bank issues a commitment letter that restates the terms of the application it is called a conforming commitment.

The other type of report about loans is the review. A review looks a lot like a presentation, but never needs a commitment letter, and has a different procedure for approval. A number of different attitudes exist in the company about reviews. Sadly, many of these attitudes are irrational. On one end of the spectrum are the people who are constantly reinventing the wheel (and the screw, the pulley, the lever, and fire). These people want to see every detail about the loan that can possibly be mustered. Since there is generally more information about a loan a year into it than when it was originally made, these reviews can get ridiculously long. And the attachments, oy. Some people put in everything but the kitchen sink, and I have the feeling that if they had a powerful enough stapler they would stick that on too. To these people I say, remember, this is a review. The loan has already been made. Certainly the bank is interested to know if the property is performing as expected but, whether it is or not, there is little that can be done about it now. We already got into bed with the borrower when we made the loan, and no amount of reviewing is going to get us out. The other end of the spectrum is less annoying, inasmuch as it creates less work for analysts, but still creates problems. These are the people who see no purpose in doing reviews. They allow loans to remain unexamined, or only cursorily examined, for years at a time. To these people I say, wake up. Doing reviews might not be the most exciting part of your job, but it is part of your job. If a loan has deteriorated, the bank should know about it, and take appropriate precautions. If a customer is doing particularly well then the bank should be pursuing that customer for more business. And if a loan is doing simply as expected, then a review of it should be a quick and painless process. At a minimum a review should contain the following: a comment on the payment history of the loan, a check of the tax escrow, a check for up-to-date property insurance and other important documents, and an inspection of the property.

Reviews are generally the first exposure to underwriting that credit analysts get. Many credit analysts become resentful that they work so often on reviews and so rarely on new loans. To these analysts I say, enjoy it while you can. Reviews are generally graded on a much lower standard than presentations. And they can afford more of an opportunity for creativity, since officers will be mellower about allowing unusual analysis. Once you are exposed to the hyper-criticism that can accompany Bob Glickman's reading of your presentations, you may well find yourself pining for those carefree days of boring old loan reviews.

Return to Introduction - Continue to Chapter 2: Documentation

*The word "he" is used throughout this manual as a convenience to avoid the clumsy he/she or the ungrammatical they. It is certainly not my intention to imply that women cannot be credit analysts, developers, title insurers, etc. Return to above.