Currently one of the industry’s hot topics, technology is transforming the way we do business. But as firms look to upgrade their technical capabilities and reach, there remains a strong case for continuing to invest in people to provide the essential and complementary human support. This is borne out by feedback from the intermediary community, especially from those brokers and packagers active in the specialist sector.
Over the last three years or so, all the stakeholders in the mortgage supply chain – lenders, intermediaries, insurers, valuers, estate agents and credit search bureaux – have made significant investments in their IT. Individual objectives may vary, but the overall purpose is the same – that is to simplify, streamline and speed up processes.
get more daily news here
As part of this, lenders and packagers want to make their processes more straightforward by integrating IT innovations with their own systems, particularly their online underwriting – or e-decisioning – systems. This is important because the best e-decisioning systems do a great job of making a highly complex integration of mortgage supply chain providers look straightforward to the intermediary and the end customer.
They start with integration at a system-to-system level with a sourcing engine but should also include electronic identity (e-ID) checks, an automated valuation model (AVM), credit searches for risk assessment and product selection. These systems are now becoming well established and mature so it is worth reflecting on their relative success and the alternatives available when the computer says ‘no’.
Referral processes
There is a significant tranche of mortgage business that can be processed on a straight-through transaction basis, such as prime low loan-to-value (LTV) remortgages on standard property types. But there is also a substantial amount of business that ticks most, but not all, the boxes in the e-decision process.
This type of business is often acceptable to specialist lenders but is likely to be turned away by e-decisioning systems unable to cope with its inherent risk complexities. Computers are not necessarily good at taking a commercial view or making ‘maybe’ decisions, so a hand-off for a manual review by a skilled underwriter will ensure the opportunity is not lost without giving it the attention it deserves.
But well-designed e-decisioning tools can also be flexible. The best, for instance, will indicate why an applicant falls outside criteria for a plan – such as the LTV being above the cap for the AVM – and suggest the solution, which in this case would be a physical valuation of the property. Similarly, a good system should provide brokers with comprehensive pre-submission data regarding what is needed by the lender.
download our news ticker
A lender’s sales team – its business development managers and internal sales desk – should also provide a vital link between itself and its brokers. A good sales desk should be able to deal with a whole range of enquiries from brokers while also extending flexibility in criteria terms where appropriate. It is very difficult for computer systems to accommodate such flexibility, which can be crucial to supporting a relationship.
Coexist
Some systems are undoubtedly more advanced than others. But despite the rise in technological sophistication there are few, if any, examples of mortgage lenders dispensing altogether with the people behind their processes in favour of computers. This is understandable given the nature of the components that typically comprise the mortgage assessment process.
For instance, e-ID has become very popular over the last two years and automated success rates of over 80 per cent are achievable. But lenders still operate back-up systems that include manual ‘Know-Your-Customer’ checks, supported by proof of residency and personal identification documentation.
Lenders also typically base their assessment of credit worthiness on a combination of historical indicators, such as court judgments and mortgage arrears or default data. Of these, arrears can be the most challenging to evaluate automatically. This is particularly so where a lender’s criteria assesses arrears performance by discrete periods, for example, in the last three months, or seven to 12 months ago. There are various scenarios where there will be insufficient e-data available to fulfil this assessment, so the submitting intermediary will need to provide a written mortgage reference to substantiate it.
Nevertheless, it is recognised that there are components of the mortgage process where human interaction is not necessary. A good example is the AVM. Typically, this either succeeds by producing a valuation within expectations – and with a suitably high confidence level to support the decision – or it doesn’t. Where it fails, a full valuation is the obvious alternative – which may, it is accepted, feature a degree of interpretative opinion in reaching its conclusion.
The technology trade off
Lenders’ written underwriting criteria can be both comprehensive and complex. Systemising it fully would result in an unwieldy e-decisioning process given the range of questions required. In truth, there has to be a balance between asking the essential questions to provide a decision and keeping the e-decision process down to a realistic number of pages.
For example, to include all the questions which might relate to property types in an e-decision would run to several pages. Just covering the range of different concrete types and approved concrete repair property types would run to several pages. It is also questionable whether the mortgage intermediary or the customer would know this data before they have sourced the valuation. Therefore, while LTV restrictions may apply at the specific construction type level, it may not be possible to capture all this detail in the e-decision.
catch up on the industry buzz
Improving offline solutions
So, how can lenders improve their refer mechanisms? Let us assume that the broker is using the lender’s website to undertake an online decision-in-principle. If they run into questions regarding the application, how might the hand off to the lender’s staff be supported by technology?
Firstly, lenders can consider instant message systems on their website as an alternative to e-mail. These have the merit of allowing the broker to type in a question and receive responses real-time and for the sales team to handle several of these enquiries from brokers concurrently. If the broker does call in, it is useful to have call statistics in order that the effectiveness of calls can be monitored and appropriate resource put in place for busy periods. When calls are received, a call logging database can be useful to avoid enquiries falling into a ‘black-hole’ and to ensure that agreed criteria exceptions are logged.
The challenge for lenders is that technology can be replicated by other lenders but it is the quality of the staff defining how the technology will work, and by the knowledge of their own user base in guiding intermediaries through using it, that will differentiate them in the long term.
If there were five or six major lenders with broadly equivalent technology offerings and processes – possibly because they have purchased their systems from the same IT vendor and consequently arrived at the same destination in terms of the capabilities – they would be offering brokers the same solutions. In such a scenario, the competitive nature of our business would dictate that a differentiator is found. This is where people would come into their own.
The reality is that there has already been a drift towards this ‘me too’ effect, particularly in the areas of pricing, product design, arrangement fees, and optimisation around sourcing system rankings. Lenders are beginning to look very similar in each of these categories. It is therefore highly likely that as some lenders increasingly pursue the same pure technology solutions others, perhaps supported by disaffected brokers, will continue to force their own proposition based on their commitment to technology with a human face.
