10 years of property claims

Property insurance exists to protect landlords from the seemingly unpredictable and almost always unexpected. Whether that’s a burst pipe, a damaged roof, or a break-in, the protection offered brings peace of mind and helps landlords when they need it most.

But what if claims weren’t unexpected? What if we could even the odds and predict what landlords might need to claim, when they’re most likely to claim it, and where in the country is most at risk?

Hamilton Fraser Total Landlord Insurance has analysed data from around 5,000 claims over a 10-year period to see what kind of trends and predictions can be discovered and how landlords can prepare for the worst – the results are surprising.

Chapter 1 - claims and incidents

1.1. Claim intensity over time

By plotting the various claim types over a 10-year period on a scatter plot graph, we get a fascinating insight into the most common claims and how they’ve changed over time.

Key trends:

  • 34 per cent of all claims paid out since 2008 have been for escape of water, defined as ‘When water has entered the property by the mains water supply and caused damage e.g. a burst pipe.’ This is two and a half times more claims than the next most common reason - storm damage
  • We paid out a total of £751,377 for 'escape of water' claims in 2011 - the highest in the last 10 years
  • There are noticeable spikes in areas with Victorian plumbing (see section 1.4 for more details)
  • Falling trees, collapse of walls and lightning damage claims are extremely rare
  • Flood claims are also rare and there are long periods (often of at least a year) where no claims have been made whatsoever. However, according to the Association of British Insurers (ABI), the average flood claim payout has risen in recent years to £50,000
  • Malicious damage claims have become less frequent in the last 4 years
  • Between 2015 and 2016 there was a notable absence in ingress of water claims

Click and drag the chart to zoom in.

1.2. Claims per capita

The majority of our claims come from London-based customers, but does that mean they claim the most? The bar graph below details the penetration of claims per 1,000 households in every UK region covered by Hamilton Fraser. The bars in the graph are graded to reflect the total number of claims in order to account for the population differences in each region. Looking at data per capita is all about attempting to reduce the anomalies that often result from less focused data.

Key trends:

  • 1,395 claims were paid out to Greater London - 88 per cent higher than the second highest region: South East England
  • Interestingly, despite only 23 claims being paid out in Northern Ireland in total, that equates to 74 claims per one million households. This compares to only 52 claims per one million households in Greater London
  • In most regions besides London, Northern Ireland and East England, the total amount of claims per capita are largely uniform

Click and drag the chart to zoom in.

1.3. Claim locations by incident type

The interactive heat map below reveals the spread of different types of major incident across the UK between 2008 and 2018. This provides a way for users to focus on their local area and understand the incident landscape over time in their area.

Key trends:

  • Storm claims across the UK saw a peak in 2014, with 149 claims being paid out
  • Since then, storm claim payouts have been on a decline with only 54 claims being made per year since
  • In 2010, accidental damage accounted for 5 per cent of all claims. This more than doubled within 6 years, to 11 per cent
  • Subsidence claims have been creeping up since 2014 by 3-4 claims a year (which equates to a 24 per cent increase YOY)
  • The majority of claims are grouped around London
  • There are also major clusters around cities such as Birmingham, Brighton and Cardiff, where private rentals are more common
Location map
Remove all markers

1.4. The London anomaly

It’s no great exaggeration to suggest that London has a serious problem when it comes to escaping water. In the Capital, there is a greater percentage of old buildings with plumbing that was first laid down over 150 years ago and that has been working over-capacity for decades. As a result, London’s water pipes have burst and leaked over 36,000 times in the last six years alone, according to Thames Water.

Escape of water and London plumbing

Our claims data marries up with the dire plumbing situation in London. The map below plots escape of water claims across the Capital between 2017-18. In Brixton alone, there were 27 reported escape of water incidents along Victorian plumbing lines, while Shepherd’s Bush saw 28 and Kensington 45 – amounting to 6 per cent of total claims made.

Interestingly, these incidents not only tend to cluster around the Thames, but closely match the city’s Victorian water mains, which can be seen more clearly in the map below. Here, the black lines represent London’s Victorian plumbing infrastructure, with the intensity of the blue markers relating to the frequency of water escape claims. You can read our full report here.

Thames Water, for their part, are at least attempting to defuse this ticking time bomb, but they might not be going far enough. They might have laid over 2,300 km of new plastic pipework across Central and North London and cut leaks by 36 per cent as a result, but according to Hamilton Fraser’s commercial and claims manager, Lorna Larke - “London landlords are right to wonder if this will be enough to protect their businesses”.

Key trends:

  • £1,166,643 was paid out for water-related incidents in London between 2017-18
  • Last year, 7.3 per cent of all money paid out across the UK went to Londoners' water-related problems
  • In 2016, 13.7 per cent of all money paid out across the UK went to Londoners' water-related problems – evidence the issue is improving, however slowly

Chapter 2 - seasonal fingerprints

2.1. Monthly fingerprints

The 2D density ‘fingerprint’ plots below show where in the UK claims are being filed each month, with the density of the pattern relating to the number of claims in each area. What’s interesting here is that January and December reveal a greater widespread intensity across the UK compared to other months, revealing that claims are more common in winter when pipes freeze and boilers are generally more active. July and August reveal a similar dissipation.

Key trends:

  • Claims are grouped around the South East during all months
  • London is intense at all times of the year, though this can be put down to our higher concentration of customers in the Capital
  • The Midlands also experiences greater claim intensity during January, April, August, October and December
  • Northern England is comparatively calm throughout the year, even during peak months
  • During December, claims in the Midlands are almost as intense as in London
Monthly fingerprints

2.2. Claim type seasonality

By mapping claim types from the last 10 years against the date of claim, we’re able to show what claims you’re most likely to make throughout the year. For example, we see a huge spike in fire claims in January, while escape of water claims are at their lowest in April. The scatter plot graphs below chart the frequency of trends over the course of a year and reveal that, seasonally speaking, there is a distinct behavioural narrative at play with most claim types.

Key trends:

  • Escape of water claims increase in the autumn/winter months, when pipes freeze and begin to leak or burst
  • Malicious damage claims peak around summer
  • Successful break-ins peak in May, before tailing off at the start of the summer. They also appear to be almost the direct inverse of theft, which peaks towards the end of the year
  • Attempted break-ins are very high in January, but the claims quickly die down moving into summer, before starting to pick up again from September onwards. Could this be because homes are seen as more vulnerable during the winter period or opportunists looking for Christmas presents?
  • There is also a notable lull in break-ins during the summer, perhaps owing to longer daylight hours
  • Accidental damage peaks in the summer when people spend more time at home and children are away from school
  • Subsidence claims are steady throughout the year, but dovetail in December
Claim type seasonality

2.3. Claim type trends

Whilst seasonality is fascinating, what about general claims trends? How have they changed over the last 10 years? Subtracting seasonality from the data and judging the claim types on an annual basis reveals that the volume of certain types of claims appear to be very stable. The scatter plot graphs below are similar in function to the graphs used above, but offer significantly different patterns when we track claim types on a biannual basis instead of by month or by season.

Key trends:

  • Theft incidents have been rapidly decreasing since 2011
  • Subsidence, fire and ingress of water are the only claim types on an upward trajectory heading into 2019, after all 3 dipped around 2015-17
  • Declining trends: theft, underground service, liability and accidental damage
  • Increasing trends: malicious damage, subsidence, fire, break-ins
  • Stable trends: escape of water, storm damage
Claim type trends

Chapter 3 - weather patterns

3.1. Storm trends

While we might not experience storms as frequently as some other western countries, major (named) UK storms do significant damage to homes every year. The graph below plots each major UK storm from December 2015 to December 2018 (the extent of publicly available storm data from the Met Office), and correlates them with the claims made as a result of storm damage in that period.

Key trends:

  • There have been 29 major UK storms worthy of being named since November 2015
  • 51.2 per cent of all storm incident payouts have been for roof damage
  • 20.8 per cent of storm payouts have been for UK hurricanes. (When high winds reach over 74mph, it’s then classed as a hurricane whereas wind speeds that are below this are still classed as a storm.)
  • December 2015 was a particularly bad season for storms, but there were also groups of storms between December 2016 and March 2017 and between September 2017 and March 2018

Click and drag the chart to zoom in.

3.2. Predicting claims using weather patterns

Aligning our claims data with weather data means we can predict with great accuracy the likelihood of storm claims in the future, giving you a chance to take preventative measures for roofs and fences, and in gardens.

The chart below reveals how closely maximum gust speeds correlate with our storm claims, with the pink line representing gust speed and the navy line representing actual claims over an 18-month period between June 2017 and November 2018.

Click and drag the chart to zoom in.

Given that weather forecast data is publicly available, we have built powerful predictive models that will reveal periods of high storm claims. For instance, an average maximum gust speed of 30 km/h over the next six months would mean an average of 14 claims for storm incidents.

Decrease the gust speed to 10 km/h and that becomes six claims. Increase it to 50 km/h, however, and you’re looking at 28 claims.

Key trends:

  • As gust speeds increase, the amount Hamilton Fraser payout increases in a reliably linear fashion
  • In January 2018 and September 2018, wind speed and claims peaked at 30kmh and 6 claims
  • In September/October 2017 and July/August 2018 there were no high winds and, as a result, no claims

Click and drag the chart to zoom in.

Conclusion for landlords

The regional data may look like it points to a volatile situation in London - the majority of our claims come from the Capital after all. But digging deeper tells a different story. Our claims per capita data actually shows it’s the least at risk for insurance claims across the whole of the UK. A close second is the East Midlands while the South East follows up in third. The long and the short of it; landlords should be mindful about the properties and areas they choose to invest in.

London does have its issues, however, particularly when it comes to escape of water claims. In 2018, 7.3 per cent of all money paid out across the UK went to Londoners' water-related problems. The city sits on a bed of clay soil that has a habit of attacking old pipes which means water damage claims are being made throughout the year, not just in cold weather or during storms. Again, when investing, landlords need to be wary of the plumbing infrastructure in the local area. If it’s Victorian, you may want to think twice.

Of course, the issue isn’t confined to London; escape of water has been, and continues to be, the most pressing concern for landlords across the UK. This commonly occurs due to cracked or damaged pipes, which often feel the brunt of the winter chill.

The fascinating thing is how seasonal claims are, not just the obvious things like storm or water damage, but also claims like break-ins, accidental damage and subsidence. This historical data points to trends which should make landlords sit up and pay attention. Break-ins, for instance, see significant spikes in summer and winter months. Could a small note to tenants around this time on security best practices lessen the risk?

Similarly, fires spike significantly in winter. If tenants are leaving the property unattended over winter and leave power on throughout the house, this increases risks. Landlords might want to find out their tenants’ Christmas plans, check the property over this period or, again, educate them on best practice on how to avoid potential problems. It might also be worthwhile stipulating a maximum period a tenant can leave the property vacant, e.g. two weeks. No-one wants to return from Christmas celebrations to a fire damaged house.

As insurers, we always recommend vigilance in maintaining your property, keeping it in the best condition to weather any storm (metaphorically and literally). Prevention is always preferable to cure, but our data clearly shows that investment in maintenance can be seasonal to match claims. If you want to invest in roofing, for instance, do so before the spike in storm claims. If you want to check for subsidence, do so in summer.

It’s impossible to know for certain what the future may bring, and insurance is there to protect you in any eventuality. But using the data above means you can think and work smarter to protect against huge problems.

About the data

The dataset used for this report is made up of 4,890 Hamilton Fraser claims records filed between 2008 and 2018 across the breadth of the UK. This data was blended with: (1) risk ratings for each postcode region; (2) UK population and household data; (3) geographic data that assigns postcodes with a latitude and longitude; and (4) higher-level UK region assignments for each postcode.