Mortgage fraud is on the rise and fast becoming a problem for lenders. Stephen Gallagher and Graham Ure look at the drivers behind the emergence of mortgage fraud
It is, by now, common knowledge that the origins of the credit squeeze, which is being increasingly felt by institutions and individuals across the world, lie in large part with lending problems in the
A combination of seemingly endless house price rises, increasingly easy credit conditions, packaged and re-packaged loan portfolios securitised and traded on the capital markets, and stream-lined application processes (such as self-certification) provided a false sense of security where profit took centre stage over risk. There was less scrutiny of borrowers’ real affordability, and more scope (not to mention financial incentive) for unscrupulous parties to engage in fraudulent behaviour, both on a private individual and an organised group basis, for personal gain. As interest rates rose, the housing market cooled and prices fell, this apparently rosy situation started to unravel.
So, how big is the fall-out? The Association of Chief Police Officers (ACPO) estimates that mortgage fraud costs approximately £700 million per year; although this figure is open to debate, there can be little doubt that it represents a major concern for lenders, especially during a period where collective belts are being tightened and maximising liquidity is the order of the day.
The Financial Services Authority has now officially entered into the debate. So far in 2008, it has banned or fined 17 mortgage brokers who have been implicated in making actual or potentially fraudulent applications. It has also recently announced that it is targeting 200 further intermediaries to ensure they have sufficient checks in place, and is developing a database of mortgage fraud patterns. The FSA clearly both expects lenders to tighten up their defences against mortgage fraud, and more importantly believes some may not be guarding themselves with sufficient diligence. As Philip Robinson, head of financial crime and intelligence at the FSA, stated:
"Lenders must also have in place systems and controls to identify and reduce fraud, and continue to provide us with the intelligence which is key to success in this area."
With the topic of mortgage fraud firmly on the radar of the regulator as well as the British Bankers Association and Council of Mortgage Lenders, and no sign of the turbulent economic conditions abating anytime soon, lenders have little choice but to examine closely what they are currently doing to prevent, detect and take action against this activity. However, where to start? What are the common types of fraud being committed against them, and by whom? And, most importantly of all, how do they begin to find the bad loan needles in the portfolio haystack?
Defining mortgage fraud and why it is an issue
Mortgage fraud is defined by the FBI as the “intentional misstatement, misrepresentation, or omission by an applicant or other interested parties, relied on by a lender or underwriter to provide funding for, to purchase, or to insure a mortgage loan”. In simpler terms a mortgage fraud is a transaction in which misrepresentations or omissions occur from which one or more persons benefit or profit.
The subject of mortgage fraud poses real concern for mortgage lenders. Application fraud is increasingly an issue now that we are starting to see significant depreciation in the value of property with no established view on how low the market will go. The spectre of managing defaulting mortgage payments and managing arrears looms larger.
Organised crime rings are more of an unknown quantity. The Thamesmead horror stories have spread through the industry giving rise to lending black-spots in East London and
In both instances there are two key issues for mortgage lenders:
KPMG’s Fraud Barometer identifies that in the first half of 2008, there were nine cases of mortgage fraud (with a value of £100,000 or more) in the courts with a combined value of £20.8m. This compares to just 10 cases valued at £3.7m for the whole of 2007. This may well prove the tip of the iceberg as the credit crunch continues to unfold.
How then can both historic and future mortgage fraud be detected and prevented? To establish this, it is necessary to identify fraud types, and the principal characteristics of fraud. Developing a profile of the fraudster is a critical first step. The Association of Chief Police Officers report states that mortgage fraud was generally committed by men in their mid- to late-30s, with nearly half of cases (46 per cent) found in London. This doesn’t go a long way to detecting mortgage fraudsters – how many of us reading this article now, fall into this broad classification?
Categorisation of mortgage fraud
Mortgage fraud can be broken down into two key types:
Fraud for Housing: This type of fraud is committed by individual mortgage applicants who give false information on their mortgage application form, either about their income or other financial obligations, in order to get their loan approved.
Fraud for Profit: This occurs where organised gangs or individuals collude together to obtain mortgages fraudulently in order to make profits or to use them to launder the proceeds of crime from other criminal activities.
Let’s take each in turn and define the nature of the fraud in a little more detail and provide examples of how the fraud is perpetrated. In this context we will consider how technology and specifically data analytics could be applied to assist in the identification and potential prevention of the outlined mortgage fraud.
Reducing Fraud for Housing through application of data analytics
This type of fraud most commonly occurs where buyers are able to declare an income without verification – often referred to as “liar’s loans”. Over the past few years, relaxed borrowing conditions enabled mortgage applicants to take out self-certification mortgages where lenders have not required proof of income to be provided, greatly increasing the opportunities for mis-statement. This situation can also occur where the lender has not concluded adequate checks on the information provided or where it has been falsified. Borrowers may also fail to disclose or conceal other obligations, such as mortgage loans on other properties, or newly acquired credit card debt.
As an example of this type of fraud, a recent case of Fraud for Housing resulted in an individual being jailed for 12 months at Portsmouth Crown Court. In this case, the fraudster took out mortgages on five properties in the Havant area totalling just over £450,000, on the basis that he was in fixed employment earning up to £46,000. In fact the fraudster was unemployed. In this case the fraudster had been provided with a bogus employment reference which enabled the securing of the mortgages. This is an extreme example of applicant fraud.
Technology is capable of detecting and preventing this type of fraud and ensuring that it does not degrade the quality of a loan book. A first line of defence is the National Hunter System and its credit scoring capabilities. Application of this solution should highlight applicants with poor credit histories.
Potential fraudsters will often attempt to ‘beat the system’ by falsifying address details. Attempted mortgage fraud can be detected by reviewing loans and loans files against known fraud indicators. The use of rules-based analysis can be applied to run potential loan transactions against known fraud schemes and indicators, known as red-flags. In this instance, a rules-based analysis might have identified discrepancies regarding address history, an over-stated income against expectations set by occupational description, and an anomalous pattern of home ownership. These red flags exist as a set of rules applied to loan application data. Suspicious loans which trigger the red flags can then be investigated whereupon in this case the lack of P60 and salary documentation would identify the falsification.
Rules should be constantly updated as fraudsters are continually testing the boundaries; fraud is like a balloon – squeeze it in one place and it will inevitably pop up somewhere else.
Reducing Fraud for Profit through application of data analytics
In December 2007, the Law Society sounded the alarm that organised criminals are increasingly focusing their activities on mortgage fraud, concentrating in particular on the buy-to-let market. Examples of fraud “ring” activity have included the overvaluation of new-build properties with the collusion of criminals and members of the mortgage application chain including brokers, surveyors and solicitors.
A hypothetical example of this type of fraud is where a criminal network arranges to buy a property from a developer for £200,000. The criminal network then identifies a nominee to take out the mortgage on their behalf and with the help of a corrupt surveyor and solicitor they arrange a mortgage for £300,000 by using self-certification and targeting sub-prime lenders. On the day of completion, the solicitor “flips” on the property, with the innocent developer being paid the £200,000 and the balance going to the fraud ring as “profit”.
For a period of time the mortgage is paid using tenants. In the meantime, additional mortgages can be taken out until on a given day the fraud ring disappears having systematically defrauded the mortgage lender who is left with the arrears on a property valued at significantly less than they had been led to believe.
To date the levels of fraud for profit have not been quantified. This is in no small way due to the fact that identifying collusion is difficult. Traditional rules and models based analytical techniques have been developed to identify individual instances where an applicant is exhibiting suspicious behaviour. A rules-based approach might contain a red-flag where there has been a significant increase in the valuation of a property, but this does not in itself constitute an open and shut case of fraud.
However, new analytical techniques for the identification of collusive behaviour are available, and can be applied to addressing the dual issues of determining current levels of exposure to fraud on the loan book and the ongoing requirement to detect and prevent mortgage fraud. Traditional analytical techniques can be combined with the application of social network analysis which identifies new instances of previously undetected fraud by linking together connections (across multiple data-sets) and finds relationships within the data. Organised criminal activity and fraud rings can be identified through understanding these relationships, and suspicious cases can then be prioritised for investigation.
In the case of the example, social network analysis could be used to identify groups of loans or events linked through relationships to other entities, e.g. brokers, solicitors and surveyors, both historical and current. The key to detecting collusive fraud is to be able to analyse all relationships and activities connected to a loan or series of loans. These networks would then be analysed and scored against known fraud techniques and patterns which are indicative of fraud; these might have included surveyors providing valuations outside of their normal sphere of operation, individuals with obfuscated names and address details with a greater number of buy-to-properties then previously recognised. In each of these cases this information could have proved the trigger for the loan to be turned down or investigated further.
Addressing the challenges
After nearly a decade of easy lending conditions, the credit crunch has brought us to the point where the metaphorical tide has now gone out and we will see who wasn’t wearing shorts - ascertaining existing levels of fraud and mortgage arrears risk is a key concern for the lending community. We believe that the increased use of data analytics, combined with traditional approaches of fraud risk assessment, can be combined to better quantify fraud loss and determine where those bad loan needles are.
Stephen Gallagher is a partner and Graham Ure is senior manager in KPMG Forensic
Date: 4th, September, 2008
ADVICE TO READERS
While this website is checked for accuracy, we are not liable for any incorrect information included. We recommend that you make enquiries based on your own circumstances and, if necessary, take professional advice before entering into transactions.
Enter your postcode here to find out how much your property is worth, based on Land Registry data.
RSS Feed
Delicious
Digg
reddit
Facebook
StumbleUpon.jpg)