Unwrapping 2019

Having just exited 2018, and as we step blinking into 2019, the team at Houseprice.AI thought that we should take a look at the year ahead. So to start with, we want to talk about property predictions and their accuracy.


As the Chinese proverb goes, the appropriately named year of the Pig promises to be an interesting time and clearly not just for the property market. All, largely due, to the ever surprising package we call Brexit.
So even though all the presents of 2018 are now unwrapped, one is sitting there and as we give the box a squeeze, we still have no clue what this one marked 'Brexit' will actually look like because the tag still reads 'Only open on March 29th'.

To help us we have reviewed a number of possible outcomes and scenarios and we have combined a number of resources and estimates from the following sources: Bank of England (BOE), Office for Budget Responsibility (OBR), Institute for Fiscal Studies (IFS), Centre for European Reform (CER) , Centre for Economic Policy Research (CEPR) all of which offer current best estimates for the UK’s economy and either explicitly or implicitly, associated potential impacts on the UK's residential property market.

Now many readers will say, 'ahah here's yet another crystal ball, mystic Meg exercise'. However, prediction is actually our business, and we are pretty good at it when it comes to property. By using objective data, paired with Machine Learning (ML), all expressed within a probabilistic framework, we are able to estimate property values to average errors of 2-3%. Our approach is scientific, by this we mean it is expressed as how probably wrong we are, not how absolutely right we are.

However there are three points that need to be stressed.

Firstly, it must be stated that the following are scenarios not forecasts. The scenarios illustrate what could happen, not even necessarily what is most likely to happen under a set of key assumptions. It is therefore a spectrum of outcomes that we worked with.


Secondly, whilst Brexit is clearly the major domestic factor for the UK, the global macro-economic environment is clearly another major concern. We could obviously mention the current fall in global stock markets, the prospect of further trade tensions between the US and China, the start of Quantitative Tightening (QT) and the flattening and inversion of the yield curve. All of the above point to much increased economic downside risks in addition to Brexit. For example, the chart below is reason for concern, since flattening of the yield curve, historically, is more often than not, a pretty good predictor of housing property declines.

Thirdly, whilst real estate is one of the largest asset classes, it is also the most heterogeneous. By comparison, traded stocks, bonds and commodities whilst also very diverse are far more standardised, so a further caveat we must make is that individual properties values can vary very greatly within a single location. Using our ML based algorithms we have high confidence for the predictive accuracy of our appraisal predictions at both a micro and property specific level. The image below links to an interactive map and this example in Stevenage - postcode outer SG4 - shows that HP.AI's average prediction error was 1.10% for 93 Terrace properties that sold and that we predicted over the previous 12 trailing months.

It should be mentioned that this is just one sample postcode area randomly taken from over 2,100 postcode areas that we estimate precise individual property values for, every month, based on our estimates vs actual Land Registry transactions. In fact, if you click on the map image you can see the sample for 12 months up to November 2018 has over 530,000 property valuations. Incidentally why only 530,000? Well we have to calibrate our ML models on the others, so some 700,000 in the last 12 months along with another 15 million from previous years.

Click the image to go to the reactive map

So whilst we are probably wrong, but we are also probably more likely to be less wrong than someone who is authoritatively pointing a finger in the air and citing subjective and anecdotal evidence. The interactive map below shows how accurate our valuations are in every postcode in England & Wales.

Click the image to go to the reactive map
We hope you have a good week!

Eldred Buck is CEO and Co-Founder of Houseprice.AI Ltd. He has over 25 years experience in capital markets and banking, specialising in quantitative models and derivatives trading across all major asset classes. Previously he founded Eiger Trading Advisors, a leading fintech company.
Need More Information?

Please contact us if you have questions about Houseprice.AI , our AI data analytics app, want access to our API, or would like to schedule a demo.

* info@houseprice.ai


Message Us

Eldred Buck interview, RICS Modus magazine

Eldred has been featured in the October issue of the Royal Institute of Chartered Surveyors monthly publication, Modus.

The article came out on the 10th, so it is "fresh off the press". You can read it here, or click the link to download a pdf copy.


Example fallback content: This browser does not support PDFs. Please download the PDF to view it: Modus Article.

Eldred Buck is CEO and Co-Founder of Houseprice.AI Ltd and a Non Executive Director at Sequant Capital. He has over 25 years experience in capital markets and banking, specialising in quantitative models and derivatives trading across all major asset classes. Previously he founded Eiger Trading Advisors, a leading fintech company.



If you would like to know more information about Houseprice.AI , Horizon, or access to our proprietary API please feel free to contact us at info@houseprice.ai.


Message Us

Introducing Houseprice.AI Data Analytics

Everything we do, we believe, challenges the status quo. We believe in thinking differently. Transparency, clarity and consistency are the cornerstones of our philosophy that drive the Houseprice.AI approach.

Our new report is designed to provide property professionals across the residential property spectrum with an innovative and accessible set of benchmarks and reporting templates.
Houseprice.AI’s proprietary reports and benchmarks provide fast and objective comparative analyses between residential projects, from single family homes through to large scale developments. This approach improves estimates of current and future sales prices, Gross Development Value (GDV) estimates as well as identifying risk with far greater precision.
The Houseprice.AI risk/reward index and Classification table further enhances residential real estate data knowledge and improve decision making and profitability.
Click here to see the example report. This is the kind of report you can create using our api. We can also produce this report, on as many areas as you wish, as part of the subscription to our app.

Need More Information?

Please contact us if you have questions about Houseprice.AI , our AI data analytics app, want access to our API, or would like to schedule a demo.

* info@houseprice.ai


Message Us

A message to all Stakeholders

In light of the recent headlines concerning Cambridge Analytica and Facebook, we would like to make the following clear to all of our clients and any other stakeholders in Houseprice AI.


We, like many new fintech/proptech firms, use Big Data, Social Media and AI as a major part of our business activity and of the services we offer. So even if all of our data sets are always anonymised and encrypted and we obviously already follow all the General Data Protection Regulation GDPR which takes effect on 25 May 2018, we additionally have one basic rule. We never do anything in business that in turn, we would not want done to us, as private individuals ourselves. It’s a simple rule and powerful, but treating others as you yourself would wish to be treated, is effective and common sense. Some call this business ethics, we would call it a fundamental right of all us.


If you have any concerns about how your data is being used, please feel free to contact us on support@houseprice.ai at any time.

Many thanks

Eldred and the Team at Houseprice.AI

Eldred Buck is CEO and Co-Founder of Houseprice.AI Ltd and a Non Executive Director at Sequant Capital. He has over 25 years experience in capital markets and banking, specialising in quantitative models and derivatives trading across all major asset classes. Previously he founded Eiger Trading Advisors, a leading fintech company.



If you would like to know more information about Houseprice.AI , Horizon, or access to our API please feel free to contact us at info@houseprice.ai.


Message Us

Bravish New World

AI. Helping you see beyond the Big Red Book


Aldous Huxley probably didn't have financial regulation in mind when he wrote his novel about a dystopian world order, but the slew of newly formed global rules, that today govern international finance, does have a touch of that Brave New World feeling.


Yes, a lot has been happening in the world of Finance lately.
The Bank for International Settlements (BIS) Basel III regulatory framework was finalised last December, with a staged implementation phase over the next 5 years; 10 if you include the Tier 1 capital ratio buffers.


Five years might sound a lot, but when you know that the best banks will want to adopt these new standards well ahead of their competition and the BIS's own deadlines, you know the race is well and truly on. Additionally this January, both the International Financial Reporting Standard 9 (IFRS9) and MiFID 2 went live too.


A cynic might argue that ultimately all of these new regulations have been brought in to ensure that the next financial crisis is not like the last one. There is more than a grain of truth in that. However, let's not forget that the 2007/2008 Global Financial Crisis's Pudding Lane, was the US sub-prime mortgage market.

Ten years on and now all of this may seem like distant thunder to participants in real estate, but actually a great part of what all these new financial fire break regulations do, is put an intense focus on Pudding Lane and particularly on property based finance and securitization.


Amongst the myriad effects designed to improve financial stability through Basel III, these new standards demand regular and better stress testing of the left hand side of the lenders’ balance sheet and specifically Loan To Value (LTV) bands and their associated Risk Weightings.

So for Assets, think real estate, commercial and residential property valuations and any lending based on these, mortgages, MBS and RMBS. These are in addition to the more obvious aspects of credit risk analysis of borrowers and credit default probabilities, along with forecasting and stress testing of future risks and then provisioning for them. All together, quite tricky stuff.

World Keeps Spinning

Meanwhile the real world has not quite stopped while these new regulatory frameworks were being figured out, let alone implemented. During this time it was not surprising that traditional mortgage lending remained and continues to remain subdued, whilst these participants have their financial probity medicine administered. Equally unsurprising that while this happened, a host of 'alternative finance' new entrants have entered into the property lending space. Now what is interesting, is that obviously these new lenders come at a cost and that actually any non-bank or non-regulated lender will likely have a much higher cost of capital, all of which will be passed on to the borrower. This is instructive, as it shows that the actual cost of a mortgage works out as the sum of the credit worthiness of the lender and the borrower plus the risk free rate. However what both traditional lenders and new participants all need, is a clear objective estimate of the collateral (underlying property), the fair value, which actually leads to the mortgage offer and thus the LTV.


Quickly you can see there is a problem, the Pudding lane problem. The GFC fire started with the tinder of poor real estate valuations, the oxygen of leverage through securitization and fanned with the accelerant of fraudulent lending criteria. We cant do anything about the last two, but we can get better valuations thanks to applying AI, Machine Learning and Big data.

Fortune favours the bravish

Over the next few weeks, the team here at Houseprice.AI will go over how several of these factors can be better measured and modelled using ML and Big Data with accurate, objective property valuations. Not exactly heroic of course, but bravish enough to manage this new regulatory maze. Our Property experts will go through how the checklist of the RICS Red Book can be better supported with objective datasets from a variety of drivers, leading to better property valuations. Our Finance and Asset Management experts will show how more accurate portfolio analytics, pricing and granularity directly affects the cost of funds and risk weighted returns for the lender. That then in turn leads directly to an end-users cost for funding property acquisitions, through a variety of distribution channels.

Eldred Buck is CEO and Co-Founder of Houseprice.AI Ltd and a Non Executive Director at Sequant Capital. He has over 25 years experience in capital markets and banking, specialising in quantitative models and derivatives trading across all major asset classes. Previously he founded Eiger Trading Advisors, a leading fintech company.



If you would like to know more information about Houseprice.AI , Horizon, or access to our API please feel free to contact us at info@houseprice.ai.


Message Us

The best UK cities for Trick or Treating 2017

In his Sunday Proptech Review, James Dearsley threw down the gauntlet for a UK company to take up an interesting challenge. Every year the American on line real estate data company, Zillow, use their analytics to create a light hearted look at the best cities in which to Trick or Treat.

So being a team that never shirk from a challenge, we at Houseprice.AI, have just crunched the numbers to find the best cities in the UK for your little monsters to score the most treats. Similar to the US based index, these cities have a population of at least 500,000 souls and are based on median home sales values, and favour those locations where the demographics show there are more children aged 10 and under and where there is high residential property density, meaning less walking from door-to-door. Lastly, because this is the country that brought Harry Potter to the world and is a global leader in potions and spells, we have added a little extra to the index, in the form of Gross Disposable Household Income. After all, you obviously would want to be knocking on doors where sweets are most likely to be lurking!

Check out the complete Trick-or-Treat Index, and the top areas in each city, below. Click on the map to see how your local area measures up on Trick-or-Treat Index!

Name Population Child Rating Median Price TT Index
Birmingham 2440986 96.85% 165k 67.96
Leeds 1777934 82.56% 162k 58.95
London 9787426 89.31% 495k 53.76
Bristol 617280 82.59% 245k 49.99
Manchester 2553379 88.72% 142k 47.55
Sheffield 685368 76.49% 130k 46.88
Leicester 508916 93.30% 175k 43.52
Liverpool 864122 73.54% 127k 41.82
Newcastle 774891 74.35% 132k 39.08
Southampton 855569 78.97% 260k 39.01
Nottingham 729977 81.32% 142k 38.17

The Houseprice AI Trick-or-Treat index is made up three equally weighted variables, the median sales price, the Regional gross disposable household income (GDHI) by each Local Authority and the proportion of all children up to and including 10 years of age as a percentage of the whole population by each Local Authority. Each of these variables is compared to the highest value nationally to create the component variable. The index is based on a sample of 149,783 property transactions from all cities and urban groupings in the UK with a population of over 500,000.

Happy Halloween from the team at Houseprice.AI!

If you would like to know more information about Houseprice.AI , Horizon, or access to our API please feel free to contact us at info@houseprice.ai.


Message Us

Uber Alleys

The news on Friday that Transport for London (TfL) had denied Uber an Operator Licence for the capital has stirred up some intense feelings. TfL announced Uber “not fit and proper” to hold a private hire licence and that the company had shown a “lack of corporate responsibility” in relation to public safety. As soon as the judgment was announced the company stated its intention to challenge the ruling in the courts and a petition was launched which is now well on the way to reaching 800,000 signatures. TfL are now saying Uber needs to rethink its approach.

Uber claim that 3.5 million Londoners rely on them to get around London, so we thought that we would look to see exactly where these Londoners are using the App. To do this we used Uber’s own API to get waiting times from 1000 randomly selected locations across Greater London that were within a 50 km radius of Charing Cross and cross referenced it against some known property metrics.

Uber Response Rates

Please click the map above to view our reactive study maps. You can click on the maps to reveal information, and double click on them and select areas.

So what did we find? Well overall across the random sample of 1000 covering the most of Greater London, the average response time was 432 seconds, or 7 min 12 seconds, however the attached visualization shows that much like property values, response times lengthened dramatically for a few postcodes the further away from Central London you travelled. This is not so surprising, you would expect more Ubers to be located in the busy central areas.

Please click the map to view our reactive study maps

However what is interesting is that actual response rates were very similar over a much larger area of London than you would expect and pretty fast too, with 2 to 3 minutes not unusual. The second chart shows the average price per square meter for properties in the same postcode areas. So if Uber were just used in central areas, you might expect the same distribution of response times as say, property prices/housing density, but as you can see, it is clearly not. In fact, response times in Outer London are very similar to Inner London and spread quite evenly, suggesting Ubers really are used by Londoners at pretty consistent levels all across the capital.
Uber Response Times Central London vs Outer London Please click the chart above to view our reactive study maps

As the chart above shows, even if you live in outer London (Green) you can have response times that are as fast as the most well served and expensive central areas (Red). So it would be very fair to say that if you are living in the most exclusive part of London, or indeed the cheapest bargain basement, in both cases your access to Uber in London's streets and alleys is about the same. That makes it a egalitarian mode of transport. This, all too often, is not the same for other modes of transportation across the capital that serve Londoners and suggests that TfL will need to rethink as much as Uber will have to smarten up its operations.

If you would like to see where the location you live ranks in London for Uber, or how other areas compare then please feel free to try out this link for our interactive charts on transportation and London property .

An extract of this blog post was written for http://www.jamesdearsley.co.uk/

Commutes, Greenspace, Parks and how to grow your own decision tree

We are all well aware if you ask any decent estate agent they will pride themselves on knowing every local detail of their stomping ground, from amenities to zoos. They will also know about transportation, schools, hospitals, council tax bands, shopping and parks, where they are and be able to rate how good they are. However if you were to ask them exactly how much each of these characteristics are actually worth in pounds per square foot, then they would probably quite rightly reply that it was all in the eye of the beholder, since buyers all have very different motivations.

So while we all know that public resources such as transportation, hospitals, parks, and schools are very important factors in housing prices, we do not have an attribution to value of what these factors actually represent. However we do know that, in aggregate, if enough buyers rate, say, proximity to a school, as sufficiently important a factor, or proximity to transportation, this will obviously lead to such properties being more or less desirable and hence more or less expensive.

Re-engineering The Decision Tree

Of course if all buyers had exactly the same motivations, then describing how these variables influenced price would be a trivial exercise. Furthermore in this respect, property portals do not help at all, since the decision tree to selecting a property is so restrictive, effectively forcing all buyers into the same space, with the same decision tree. For example, the Zoopla/Rightmove search order goes a) define location b) set the price band c) choose between flat, terrace, semi or detached house d) define number of bedrooms. To see how really unhelpful this is, imagine if these property portals were selling paintings, the listings would appears as follows: ‘Rectangle landscape, size 120 sqft, blue and green with lilies and water’, or ‘sought after portrait, size 1 sqft, mostly dark with funny smile’. Well that's a Claude Monet and a Leonardo da Vinci marketed, job done!

What if buying decisions could be based on a person’s actual unique motivations? This is where Big Data and Artificial Intelligence (AI) come in. To see why, think when people say it’s all about location, location, location, what do they really mean? Actually this translates to all those things that make up a place, which will be everything you can measure, from shopping footfall, to reported crime, schools, health, sports clubs, aspect, environment, air quality, noise pollution, transportation and so on. Indeed everything that makes a place. So far we have identified around fifty different objective measures and while it very unlikely that a buyer will list all 50 of these, they may well have a very different say, top four or five?

For an example, consider a buyer who was interested in flats with a maximum commute time to central London of 45 mins, green space, transportation and wanted to be close to a park, where would be the places they could look to buy in London?

To date, several studies on property value have largely concentrated on transportation effects and only a few studies have focused on the effect that green space has on property values. In these researchers have mainly focused on specific parks, for example Hyde or Green Park, within different communities rather than parks in general, to study the average impact of green space on housing prices. Using both parks and actual visible green space we have been able to quantify the effect of public resources on property value, especially green space and/or parks, using AI.

Machine Learning

To do this we took transaction price data and the structural attributes of 84,747 properties in and around London that have sold over the course of the last 18 months. These were then cross referenced with additional supporting data, which included structural attributes, location variables, and environmental variables. In this study, Inner London is defined as the 150 inner London postcodes. Outer London is classed as within 16 and 35 km of central London. While these studies are quite basic, they demonstrate the way the Machine Learning evolves to arrive at a precise value.

Chart1

Graph 1

Chart2
Graph 2

Graphs 1 and 2 show that the average percentage of green space has far more impact on price per square meter (PSQM) the closer you are to the center of London, this again is as one would expect, with green space being more highly prized the more built up a city becomes.
Chart3
Graph 3

Chart4
Graph 4

As the graphs 3 and 4 show, London property price per square meter is inversely related to commute distance, which is not a surprise, but it is also more widely dispersed the further you radiate out, representing cheaper PSQM for similar commute times.
Chart5-1
Graph 5

Chart6-1

Graph 6

Graphs 5 and 6 show the effect of being near a train or tube station. What is interesting is that if a property is too close, this reduces the price per square meter of a property, where noise disturbance becomes a factor to consider. In Outer London, up to 2 km away from a station or tube line and you can see very large differences in PSQM, which again shows the opportunities that exist in these areas.
Chart7
Graph 7

Chart8
Graph 8

Charts 7 and 8, show the considerable effect of living close to a park, in fact in Inner London, properties fall by and average of £4,264 PSQM every 1000 meters you move way from a park! This effect is lessened in Outer London, though it is still clearly a factor as graph 8 shows.

In fact if we look a just one property type, we get an even more impressive demonstration of how certain factors drive values. Here we are just looking at flats, but we have all property types listed under the Land Registry and based on commute distances up to 90 km to central London, we have also expressed house prices as the natural log of the price per square meter so now we can see the very clear correlation at work in graph 9 with an R squared of 0.72.

Chart9

Graph 9

We can see the effect of proximity to a park on flats as well in graph 10.

Chart10

Graph 10

Living in a Bubble

Graph 11

Graph 12


The last two graphs, 11 and 12 show that as we pull together all the unique drivers our buyer has chosen, we start to narrow down to the specific areas and property types that meet the criteria. In fact we have now the means to select individual properties, that meet the precise requirements the buyer seeks. We arbitrarily chose central London and 5 particular factors out of a much longer list, but with Houseprice.AI you could do this for any city in the UK and for 40+ other drivers.

By using AI and Machine Learning, we can arrive at assessing any property purely on the basis of objective values thanks to the breadth, detail and quality of our data. In fact this attention to detail actually leads to a more natural and holistic approach to finding property. So, property searches can be made far more intuitive, personalised and quantifiable, thereby allowing buyers to identify and locate properties far more efficiently and objectively and to be able to verify value for money on both an absolute and comparative basis.