Nicola Garbarino and Benjamin Guin
Policymakers have put forward proposals to ensure that banks do not underestimate long-term risks from climate change. To examine how lenders account for extreme weather, we compare matched repeat mortgage and property transactions around a severe flood event in England in 2013-14. We find that lender valuations do not ‘mark-to-market’ against local price declines. As a result valuations are biased upwards. We also show that lenders do not offset this valuation bias by adjusting interest rates or loan amounts. Overall, these results suggest that lenders do not track closely the impact of extreme weather ex-post.
Properties across the world are exposed to long-term risks of extreme weather events such as hurricanes, flooding and fires. Policymakers are increasingly concerned that financial institutions may be underestimating their exposure to climate-related risks. Yet, assessing their behaviour often requires assumptions about uncertain climate change scenarios.
In a recently published Staff Working Paper, we exploit a severe flood event to examine how lenders adjust their valuation of housing collateral using available information. The analysis focuses on the effect of a severe flooding event in England that resulted in over £1.3 billion in damages. In the winter of 2013–14, regions in the Thames catchment area and the east coast of England were hit by a combination of inland (river and surface) and coastal floods. By focusing on an actual event, rather than scenarios, we set a low bar for how lenders take climate-related risks into account.
We compare changes in lenders’ valuations (used for mortgage refinancing) against changes in sales prices for actual property transactions. If lenders ‘mark-to-market’, using available price information available at a local level, valuations should change in line with transaction prices. Instead, we find that valuations do not adjust to price declines in neighbourhoods that experience prolonged flooding. As a result, valuations are biased upwards. Also loan amounts and interest rates for mortgage refinancing remain unchanged. Increases in estimated flood risk (as opposed to actual flooding) do not appear to lead to a fall in property prices.
Relative to unaffected properties in the same geographic area (measured by the postcode district), properties in postcode units that experience prolonged flooding experience significant decreases in sales prices between 2.6% and 4.2%. Our best estimate points to a decrease of 3.3%. By contrast, these declines are not reflected in valuations for mortgage refinancing. We estimate a positive valuation bias between 2.9% and 3.2%. The net effect almost perfectly offsetting the decline in sales prices with a best estimate suggesting only a -0.4% change in valuations. In other words, valuations for refinancing for flooded properties are roughly aligned with sales prices of non-flooded properties in the same area.
Chart 1: Relative change in transaction prices and property valuations following long flood
A fall in the collateral value would result in an increase in the loan-to-value ratio. In the UK, it is the most important factor for mortgage rates and amounts. Additionally, we would expect decreases in property prices to be reflected in valuations, and result in higher interest rates and/or lower loan amounts for refinancing transactions. But since refinancing valuations appear to be unaffected by flooding or increases in flood risk, the effects on mortgage rates and loan amounts are limited.
Our results suggest that little local knowledge about flood risk is employed by lenders in valuing house prices following a flood. Lenders instead rely on house prices indexes that do not capture variation in flooding within neighbourhoods. If lenders used all available local information, changes in property valuations used for refinancing should be closely correlated with changes in actual sales prices. Chart 2 shows scatterplots of the change in property valuation vs sales price at different levels of aggregation (postcode unit, local authority, and region). The correlation between sales prices and valuations increases is low at the postcode unit level. However, it increases as we aggregate data for wider geographic units.
Chart 2: Change in property valuations versus sales price at different levels of aggregation
(a) Most granular level of aggregation: Postcode unit regions
(b) Intermediate level of granularity for aggregation: Local authority
(c) Least granular level aggregation: NUTS1 regions
Climate-related valuation bias can lead to distortions in lending quantities by relaxing credit constraints. A substantial reduction in the value of the collateral available for refinancing could force marginal borrowers to pay higher rates, use savings to make up for the difference, or potentially default on their mortgage payments. Biased collateral valuation can also sustain bank lending via their risk-based regulatory capital ratios, where collateral values are an important factor in setting regulatory requirements.
Our results indicate that the pricing of risks from extreme weather events in mortgage lending has so far been limited.
Nicola Garbarino and Benjamin Guin work in the Bank’s Policy Strategy and Implementation Prudential Policy Division.
If you want to get in touch, please email us at firstname.lastname@example.org or leave a comment below.
Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.