Analysis: Pulling back from the edge of the lending cliff

The last two crises were predictable and another debt-fuelled crash can be avoided, senior industry figures have concluded. Daniel Cunningham reports

By adopting a method to track long-term asset values, property lenders can spot the peak of the market in plenty of time and moderate their activity accordingly. That was the core finding of a steering group representing the UK’s Property Industry Alliance, which recently concluded a two-year investigation into avoiding boom and bust in real estate finance.

The initiative, instigated by 2014’s cross-industry A Vision for Real Estate Finance report into how to avoid another catastrophic crash, resulted in the recommendation that adjusted market value (AMV) is the best measure of when there is a potential lending bubble.

AMV uses historical data to monitor long-term value, and would have accurately predicted the 1989-90 and 2007-08 crashes, the authors said. AMV compares current indexed property values with a long-term trend line, adjusted for inflation and drawn in reference to historical data, to measure the extent of a potential crash at any point in the cycle. Put simply, the method weighs up whether values are rising above levels where, in previous cycles, crashes have happened.

When the UK’s Investment Property Databank index is more than 20 percent above the long-term trend – as it was most recently in Q2 2004 – the likelihood of a 35 percent or greater fall in the real value of that index within five years is very high.

The overleveraged UK commercial property market reached the end of a bull run in Q2 2007, and values dropped by 42 percent over the following two years. “The UK economy today still bears the financial fallout and scars,” the report’s authors said.

Using the AMV method, the current UK market is deemed to be 10 percent overvalued, meaning that a 34 percent risk of a 35 percent fall in values is unlikely to deter too many lenders. However, if the market becomes overvalued by 20 percent, the AMV analysis shows that the implications could be far more severe. Today’s lending market is relatively sober, with banks generally keeping loan-to-values in check and regulators vigilant, comments Rupert Clarke, the former CEO of Hermes, who chaired the research group. However, he warns, complacency can easily creep into the market.

“From 2004 to 2007, everyone was aware that the market was overvalued, but what do you do at that point? Reduce the amount of loans, which means reducing profit and bonuses while the competition carries on making money? It’s a big behavioural challenge. Unless it is hardwired into the system, there will be a lot of push-back and pressure to carry on,” says Clarke.

“The further away from the last crash we get, the more people who experienced it first-hand will have retired. Regulators are on-point and there are enough senior people around who learned from the last experience, but I’m less confident that will be the case in 10 years’ time,” he adds.

Plenty more scrutiny of the AMV method is needed, but it seems to hold up when run retrospectively in relation to previous crises and it was the most efficient of three long-term value metrics examined by the steering group.

Now that the AMV approach has been identified as an effective tool at the all-property level, what happens next? The first paper published by the PIA’s debt group is clear that a lot of work lies ahead if lenders and regulators are going to embed long-term value techniques into their risk management frameworks.

Lenders and regulators would ignore the findings “at their peril”, says Peter Cosmetatos, chief executive of the Commercial Real Estate Finance Council Europe, a backer of the paper. “Lenders might argue that an asset’s income is what really matters when providing a loan, and it is partly true. But in a three- to five-year non-amortising loan market, the market value of the property at maturity is critical, because getting your principal back depends on it.”
Regulators should use long-term value sensitively, rather than linking it to regulatory capital treatment, Cosmetatos argues.

“The regulator’s main goal over the next few years should be to ensure that as broad a range of real estate lenders as possible takes account of long-term value measures in managing cycle risk across their portfolios. But that shouldn’t be done in a prescriptive way: there should be room for diversity and judgment at the individual lender and loan level,” Cosmetatos says.

“There’s a strong case for AMV to be published regularly and used to assess cycle risk across the lending market and in the portfolios of individual lenders. But unless a methodology can be shown to be really robust at the individual property level, the focus should remain at the market and portfolio level. Regulators should not use long-term value to influence individual lending decisions.”

The paper does not suggest AMV becoming a measure adopted in loan covenants, rather a tool to influence initial lending decisions, to mitigate an over-reliance on market value-based LTV at the time of origination.

Creating a lending cliff every 15 to 20 years is harmful to the financial system, the wider economy, and it causes the lenders to make huge losses, Clarke says: “If a bank is lending at full tilt at the end of a cycle, the losses it makes will wipe out all of the profit from the rest of the cycle.”

SHARE