Ignorance is not bliss —especially when it comes to business and the bottom line. Yet day in and day out, lenders are operating in the dark and losing dollars in the process. Last month, I wrote about the most common misconceptions about AVMs, also known as automated valuation models, the most commonly used collateral evaluation tools. AVM are being used over one billion times each month, and I estimate that hundreds of millions of those billions of AVMs are being used by lenders and servicers in the mortgage industry. But even considering these gargantuan proportions, most mortgage folks are surprisingly unaware of how big an impact AVMs can have on their bottom lines.
This month I’m going to cover three of the most common ways AVMs are being used today. And for those of you concerned with your bottom lines, I’ll get into the concrete details about how lenders and servicers are saving — or wasting — thousands of dollars or more each month, simply by the way they approach AVM suitability.
One of the biggest misconceptions about AVMs is that they are inaccurate. Last month, we revealed the fact that standard error rates of AVMs are actually lower than those of traditional appraisals. While my aim is not to rank one form of evaluation above another, it is to point out that AVMs can be quite reliable and highly appropriate for lower risk activities and transactions. I’d also like to remind you that there is no benefit to using an unreliable tool, not where data and decisions are concerned. Accuracy matters.
This is a key point. When financial decisions are being made based on a reported dollar value, businesses need to ensure that the evaluation tools being used are producing the most accurate value possible. While this may seem like common sense, this is not the general practice being exercised in the mortgage industry when it comes to AVMs. Hundreds of millions of times each year, lenders and servicers are running AVMs without much thought as to the accuracy of the reported figures, and this haphazard approach is likely costing them more than they imagine.
The following are three of the most common ways that AVMs are being used by lenders and servicers today:
1) Mortgage Pre-qualifications
A lot of smart lenders are using AVMs in the pre-qualification process. As they’re taking a quick, cursory look at borrowers, they’re also taking a quick review of a property, to see if a deal is worth pursuing. AVMs are well-suited for this review. It takes just minutes to run an AVM, and if the right one is used, the results can be surprisingly accurate. However, if the AVM is not accurate, lenders are likely wasting money and missing valuable opportunities.
If an AVM comes in significantly below the property’s actual current value, lenders are missing opportunities in loans that can’t be made. Let’s say your average gross profit per loan is $4,000; even one lost customer a month can amount to nearly $50,000 in missed opportunities per year. If, however, the value comes in too high, lenders can invest valuable man-hours taking a loan through the process until an appraisal reveals that the transaction won’t go. Even at a pay rate of $25 per hour, a processor spending a mere four hours on a loan that will never fund is costing the company $100 on lost labor.
This situation might be understandable if an AVM returns a value of $200,000 and an appraisal comes in at $194,000, and it’s just enough to kill the deal. In cases like these, hopefully the loan officer informs the borrower that the value is tight and may not come in as needed. If, on the other hand, the appraised value comes in at $186,000, and there was clearly no reason to even run a full appraisal, chances are, you’re going to have a very upset borrower on your hands. Incidentally, in case you’re wondering, yes, there is a way to make sure that you have significantly fewer of those $200,000/$186,000 scenarios. That is absolutely within a lender’s control. I’ll get to the details of how in a moment.
The Cost of Losing Trust
Back to the $200,000/$186,00 situation. There’s a more significant, long-term financial issue at hand here. Borrowers traditionally cover the cost of the appraisal, which is roughly $350 for the average non-FHA loan. From what I hear, borrowers don’t take very kindly to paying fees for a loan that has no possibility of being funded. While this is important for any lender to remember, credit unions, being the member-driven organizations that they are, might want to be particularly attuned to this.
There have been numerous studies of the costs of losing trust, and consequently losing a customer who feels he or she has been wronged. Anecdotal evidence shows that a disgruntled customer will tell between eight and 16 people, with 10 percent telling more than 20. When you factor in the ease of communication and the viral nature of social media, the number increases exponentially. Even if they tell no one, 9 out of 10 unhappy customers will never purchase goods and services from you again. Not a good consequence when, according to Gartner Research, the cost of acquiring a customer is five times the rate of retaining existing ones.
2) Home Equity Lending
There are two ways that lenders use AVMs with home equity lending. The first is to value the property for a home equity loan or line of credit. The second is to evaluate the line after it has been issued to determine whether or not the property has enough value to support it. In other words, lenders often determine whether or not to increase, decrease or shut down a line of credit based on the borrower’s home equity.
Home equity lending requires that an evaluation must include a physical inspection of the subject property, but it does not require a full appraisal, so a lot of lenders opt to use AVMs in conjunction with a property inspection report. As with mortgage pre-qualifications, lenders can lose home equity business when they use AVMs that erroneously under-report the true value of a property. That can add up to losing literally thousands of dollars each year over the course of several years, since home equity lending brings a steady income stream for lenders. With variable rates that range from three or four percent to 11 or more percent, and a loan amount of $100,000, the lender stands to lose $2,000 to $10,000 or more each year, over the course of years, for every lost customer. Lose one customer a month and that number can jump up to $100,000 or more each year.
If the lender is using AVMs to evaluate its home equity portfolio, and that AVM erroneously over-values properties, that lender is at risk of losing the additional income generated by higher interest rates charged to borrowers for higher loan to value ratios. However, if an AVM erroneously under-values properties and the lender reduces or revokes a credit line, it could be losing out on the additional income from the increased line. What’s more, if a lender was negligent in selecting the AVM, and revoked a line without just cause — meaning it can’t prove how and why an AVM was deemed suitable — it puts itself at risk of not only noncompliance, but also litigation.
A Public Relations Nightmare
According to numerous news sources, including CBS, Bloomberg, ABC, and the Los Angeles Times, at least two of the country’s top lenders have been named in class action lawsuits for illegally freezing borrowers’ home equity lines, which according to some sources essentially violates part of the banks’ agreement for the bailout. Plaintiffs’ attorneys cite numerous experts who have stated that erroneous AVMs are at the root of the allegedly illegal actions.
While I am not here to deem an organization guilty or not guilty, I would like to point out the repercussions of this type of debacle. In addition to the hundreds of thousands of dollars in legal costs of defending a class action lawsuit, and the tens—if not hundreds—of millions of dollars a settlement could cost, there is the issue of the public relations nightmare that follows a disaster like this. While the big banks have legal teams and large PR firms to handle these types of issues, small shops, midsize lenders, community banks and credit unions do not. A class action lawsuit may or may not drag a lender into failure. At the very least, it can play a major role in the shortened lifespan and immediate profitability of an organization.
3) Portfolio Analysis
Lenders and servicers are required by law to analyze their portfolios at least once per year. Most servicers conduct this analysis every six months, and most use AVMs for this type of analysis because they’re so cost- and time-efficient. However, there’s more to the story. Any business entity with a servicing portfolio is required by law to hold back anywhere from 20 to 50 – or in the case of home equity loans, 100 – percent of the value of their portfolios in liquid reserves. That figure is traditionally based on values derived from AVMs.
If a company values its portfolio too high, it is unnecessarily tying up valuable liquidity. To put a dollar figure on this, let’s say that the AVM you’re using returns values that are, on average, 15 percent higher than the properties’ actual value. For a portfolio valued at $50 million, you could be tying up an additional $7.5 million of funds that you are perfectly entitled to keep liquid. On the flip side of that coin, if the AVMs you’ve selected undervalue the properties in your portfolio, you are exposing yourself to the fees and fines associated with violating Basel III requirements. Portfolio analysis is a fine line with potentially severe financial consequences on either side. Your AVMs’ accuracy can tip the scale one way or the other, to the tune of anywhere from thousands to millions of dollars.
Accuracy Testing Pitfalls
As these dollar figures indicate, it’s critical to use the most suitable AVM if you want to reduce loss, maximize revenue and maintain existing customers. Most lenders and servicers do some type of accuracy testing, but chances are, they’re not using the most cost efficient means.
Some companies put AVM selection in the hands of the individual ordering the AVM. That’s not a good idea, because it’s virtually impossible for a human to be able to analyze AVMs well enough to rank each AVM sufficiently to build the most suitable cascade for a property. Others rely on their AVM providers to advise them in building their cascades, which is probably a slightly better solution, unless your AVM provider has an interest in one or more of the solutions it provides. Still others opt to use analysis companies that conduct in-depth research and accuracy testing for each of the AVMs that the lender or servicer uses. While this is certainly a better option than a staff member or AVM provider, it can be expensive and time consuming. These reports cost $50,000 to $60,000 a piece and are generally done quarterly, which brings the annual cost to $200,000 to $240,000 or more per year. And because they’re generally produced by actual humans, you can expect a wait time of roughly six weeks.
Your best choice is to use a technology designed specifically to be flexible in determining AVM suitability, like Platinum’s OptiVal, which can determine the most suitable AVM cascade in minutes. Whichever technology you choose, make sure to use a flexible technology, one that can factor in risk factors like loan to value, type of transaction, and borrower’s credit score. Each AVM is going to function differently for different properties. There’s no such thing as a one size fits all AVM. The difference between using the most suitable AVM and the least could be vast, as can the difference between using the most and least suitable appraiser. An appraiser that specializes in Des Moines properties couldn’t produce an accurate value for a property in Brooklyn the way one who specializes in the area could.
Stop the Loss
It’s easy to underestimate the role that AVMs can play in a company’s profits and loss. With their low cost and speed, a lot of folks jump to the conclusion that AVMs couldn’t possibly have much impact on the bottom line. But they can and they do. It’s time to get out of the dark. AVMs are the mortgage industry’s most commonly used collateral valuation tools, bar none. But, used improperly, AVMs can do more harm than good. In fact, I believe AVMs are slowly, methodically draining profits for virtually every lender and servicer in our industry, to the tune of thousands of dollars each year. And that’s simply because lenders and servicers have a haphazard approach to AVM suitability and selection.
Accuracy certainly matters when using AVMs. In the mortgage industry, virtually every lending decision is hinged on collateral value. We can’t fool ourselves into thinking there are any cases where accuracy is not an issue, when in fact, inaccurate AVMs can lead to losses as subtle as missed opportunities or as dramatic as compliance violations, lawsuits and tied up liquidity. Public relations disasters and reputational risk are harder to measure, but can end up being extremely expensive for lenders and servicers.
It’s time to stop the loss and start taking AVMs seriously. Unless of course, your organization has tens of thousands of dollars to waste each year.
About The Author
Phil Huff is CEO at Platinum Data Solutions. Phil is a CEO with a history of growing companies whose technologies revolutionize manual mortgage processes. As co-founder and CEO of eLynx, Phil built the management team, grew recurring revenue to $15 million, and orchestrated the company’s sale to American Capital for $40 million in 2004, five years after the company’s launch.