A tale of two syllables
The word ‘yes’ forms a complete, grammatically correct, sentence in English. Short. Sweet. But not short enough, apparently. The word ‘no,’ while not nearly as sweet, is 33% shorter. Thus, in an attempt to be efficient, mortgage banks tend to decline many potential customers, as if getting to yes is expensive.
Since brevity is the soul of wit, and tediousness the limbs and outward flourishes, I will be brief, my liege. Working so hard to get to 'no' is mad.
Number 9: Let's try getting to know
This is the ninth in our series of articles about customised home financing packages. In this piece, we explain why, in the world of AGI, getting to know is better than getting to no. Say yes, if you concur.
I encourage you to dip into our prior articles to understand how our daysrent® home purchase plans address the twin challenges of housing affordability and suitability in sophisticated ways for applicants banks are minded to reject. Each of those articles is a long read. Re-imagining the largest market in retail financial services requires thought and thinking takes time, if you're doing it right.
Joking aside, this is a story about FinTech; though it's heavy on the Fin and light on the Tech. Our most piercing observation is this: in the very near future, banks will use AI, clever service design, and troves of data to empower customers to tailor their own services, at the point of need.
Yes. That's obvious but this second insight is not. I'm talking about manipulating abundant and cheap data using human interaction and four-function calculator maths. I'm not talking about brain-bending equations that require large language models, billions of qbits, and mainframes on-prem. In the very near future, technology will empower banks to slow things down, take more time, dumb things down, and repeat themselves, if necessary, in order to innovate quickly.
Conventional wisdom suggests mortgage banks should spend less time and money at the front end of the process
Conventional wisdom is about to change. Today, mortgage banks incur most of their operating costs finding and bringing new customers on board and dealing with payment delinquency. Performing loans cost nothing to administer. Thus, spending less time avoiding default risk, automating processes, and taking humans out of the loop is a valid, pre-2024, strategy.
Next year, forward-thinking financial institutions will do the opposite. They will start hiring like crazy to develop services that can only be delivered with the technology being developed today. I mean, like, right now. Even as we speak. Sam Altman and friends are creating a Leviathan called Archie (my nickname for AGI). Sure, Archie 1.1 is a machine, but it will be awfully difficult to tell when his clone, Archie 2.3, is born in early 2025. Things are moving blindingly quickly.
Archie 2.3 will cause an explosion of new job opportunities as banks realise that taking humans out of the loop is a poor idea. It looks like a winning strategy, today because banks offer borrowers short-term contracts that revert to punitive variable rates. This misguided pricing step-up creates a perverse incentive for customers to peruse the bank's competitors' latest offering every two to five years – a conventional, short-term, short-sighted approach that will seem quaint in hindsight a few years from now.
Shortly, with Archie's help, banks will twig that generating long-term shareholder value means generating top line growth
You can't cut costs forever and shrink your way to glory. If operating costs fall monotonically, at some point, marginal costs become asymptotic to zero, for you and your competitors. Yes, that's right, everybody will have access to Archie 2.3. All costs, in all industries, will trend downward in the new age of abundance.
In the long run, banks and other service providers – financial and otherwise – will realise they must provide better services, even if they cost more time and effort to deliver. That starts with spending more time talking to customers upfront and developing more sophisticated ways of customising services in an attempt to get to yes for applicants your competitors are minded to reject.
Since Archie doesn't take lunch breaks and happily works on weekends, he will have plenty of time to ask prospective customers what new-fangled services they're willing to pay for. Cleverly, Archie will be adept at talking in male, female, and non-binary inflected voices, regional accents, and a variety of languages. Archie will be happy to listen to customers' questions, for hours, patiently think through answers, and be responsive to their needs. With Archie's help, operating costs at the margin will fall and employment will soar at those banks which make the most creative use of this new technology to take bread out of the mouths of banks that fail to provide new services.
The value of intermediation will fall fast when it costs nothing to talk at length to prospective customers
Today, plenty of fat remains on the bacon, begging to be sliced off. In a prior article in this series, Number Four: Pin Back Your Lugholes, we noted that the typical mortgage broker charges procuration fees averaging thirty-five basis points on the value of the loan. So, a £200,000 mortgage costs the bank about £700. At £50 per hour, this implies fourteen hours of yackety yack.
Pencils at the ready:
I very much doubt that Archie costs £50 an hour so traditional mortgage brokers need to look lively. An awful lot of buggy whip manufacturers in 1905 didn't think through the brutal mathematics of exponential growth in the auto industry.
Here's another confident prediction. Given the dizzying pace of change, people will crave the sound of the human voice and may even want to talk to advisers in face to face interactions in (horror) bank branches, something Archie is not very good at doing. In the age of Spotify, summer music festivals and Taylor Swift concerts are more popular than ever. The same will be true of financial services. People will want more, not less human yackety yack. And this is where things get interesting so strap in.
The general level of financial literacy among home buyers is incredibly low
First timers in particular need advice. Lots of it, but concepts such as LTV, LTI, APR, SVR, and SDLT are alien to most people. Human mortgage brokers today sound like robots. The fact that Britain has 5,500 mortgage advisory firms that use words like 'product fees' and 'redemption penalties' and not words like 'two-bed terraced house near the Montessori School' is a solid indication that the mortgage itself needs a root and branch reappraisal. Paradigms, like tectonic plates, shift slowly, ever slowly, then boom...
At bottom, a mortgage is a means to an end; a necessary evil. People who want to finance a home are really asking for help to bring up their kids in this street over here rather than that street over there. I'd be willing to wager, they'd rather spend a few hours getting advice about shaving hundreds of pounds off their heating bills, or their council tax, annually for decades, rather than shaving a few hundred pounds off a mortgage product fee just the once. Those recurrent cost savings are a function of the choice of property, not the fiddly little features in the loan.
These highly predictable long-term changes in expenditure patterns have profound financial implications for home buyers
FinTech innovation in the home buying space needs to take location-sensitive data into account, along with a blizzard of other factors that banks routinely ignore today because ignoring data and getting to no seems cheaper than getting to know how real-world issues affect their customers. I'm not talking about sophisticated geolocation – though our platform uses that data. I'm talking about sitting down with a cup of tea and asking the applicant how old their kids are and where they'd like them to go to school. The answers to those questions are intimate bits of private data that banks today simply ignore.
As we can't solve every problem in a single article, let's break down the most important, most flawed, and (ironically) the easiest to fix component of today's broken machine.
PROBLEM: Picking through an applicant's income and expenses, hunting for reasons to say no, is time-consuming, expensive, and a strangely profligate way to treat the marketing budget for banks battling it out in a mature market.
SOLUTION: Spending pattern recognition technology is cheap, quick, and increasingly accurate. It starts with the clever manipulation of troves of data using techniques that were science fiction in 2014, that are commonplace today, and which will be damn near free in 2034, or sooner if Sam Altman gets his way.
But getting hold of the data is pointless if you don't enrich and interpret it and then use it to deliver different financial solutions. We're talking about talking to your customers. We're talking about getting to know them via their data – sifting the data, kneading it, discussing it with the customer who owns the data, slowly, patiently, and for hours, if necessary, using Emotional Intelligence, not Aspartame Intelligence. Since Archie can do all of that cheaply banks that use him to provide more yackety yack will do better than those who see him as a cost-cutting tool.
Detailed spending pattern analysis forms the ideal basis for Archie and your customer to swiftly tailor unique home purchase plans
I am all in favour of using open banking data to achieve an algorithmic assessment of the applicant's ability to pay. However, that's largely a waste of time. Risk reduction is about getting to 'no.' Boring. Pointless. People don't default on mortgages. Rounded to one decimal place, the rate of repossessions compared to the number of loans outstanding is zero. Don't look at me. Look it up. The data is open-source, abundant, and stable across the economic cycle.
Let’s ground this piece in today's commercial realities. Jeremy Hunt, Chancellor of the Exchequer and Second Lord of the Treasury, recently observed that the chronically woeful stock price performance of the large British banks is a threat to Britain's place as the putative financial capital of Europe.
A bank whose price per share languishes below its book value per share (for decades, not years) will be torn to shreds should a Meta, a Google, or a JP Morgan deign to take a serious run at their core business, just as Santander and ABN Amro have done. And didn't Apple recently snap up a British FinTech focused on spending pattern recognition? Kudos to them.
Meta, Google, Apple will put Archie to work in creative ways far faster than incumbent banks currently realise
If they want to compete in the big leagues, British banks shouldn't take these Leviathan's on by developing their own Archie. Rather they should use their home court advantage, abundant, open-source data, and reams of private, permissioned data to look for opportunities to grab a large share of a vast, local, lucrative, untrammeled space.
With the aid of a Google machine, I have uncovered the following data points:
FACT ONE: The top ten mortgage banks control about 90% of the market.
FACT TWO: The market is ex-growth and banks are price-takers, not price-makers.
FACT THREE: Rental yields dwarf mortgage rates. A five percent mortgage rate, compounded over 30 years, is far less than a five percent rental yield, rising 2% faster than inflation, compounded over sixty years. Check the maths. It's not hard and the comparison is not close.
POLITE SUGGESTION: Banks should look out the window at the many millions of renters who would rather be owners and think hard about ways to get to yes. This starts with getting to know them.
Think about it this way. Roughly ten million households rent. Annually, fewer than four hundred thousand buy their first home with a mortgage. That’s less than 4%
The letter A constitutes roughly 4% of the alphabet. The letter A is first among equals and forms a complete word all on its own. However, coherent speech requires creative use of the letters B through Z. Since most renters would rather own, helping the remaining 96% ought to be tempting for a moribund mortgage bank.
Banks that refuse to use every letter in the alphabet will hand their most lucrative opportunities to Alphabet
Yes. I know. Many renters are not "mortgage ready" but what does that really mean and, even if true, why does it take so long to reach that conclusion? It can take anywhere between several days to several weeks for a bank to confirm a mortgage offer. With apologies to Marvin Gaye, what's going on? Why should it take so long to establish the applicant's ability to pay? It only takes a few minutes for a lettings agent, armed with 2024 technology, to pre-qualify a tenant.
Armed with Archie 2.3 that few minutes will rise, not fall. Why? Credit assessment will become easier, leaving Archie time to talk to applicants about many more interesting things than the risk of default. Archie doesn't care about spending time. He has all day but he needs to be told to talk about interesting things that are relevant to the customer; things that open the door to services the customer will readily pay for.
To break this down, let's revisit mortgage repayment risk, deposit protection, and other 19th-century fears that float around in an evidence-free zone. I was chatting to GPT3 the other day and, unprompted she volunteered a twelve-point list of things renters should do to demonstrate mortgage readiness. The insights were startling: Save for a deposit. Maintain steady employment or income for a consistent period. Maintain a good credit score. Doubtless, that's good advice. However, way down at number four, she recommends:
Recommended by LinkedIn
Ensure your income comfortably covers the mortgage repayments
Hold on. Isn’t that the whole ball game? If your income exceeds the amount you feel comfortable paying what else do you need to know?
Our daysrent® purchase planning process starts when we ask just two questions. Where would you like to live and how much do you feel comfortable paying? No field on the first page of our application form (see screenshot below) requires you to tell us if your parents are flush with cash or how much of your life savings you're willing to punt on a home buying journey that entails the risk of negative equity.
We show homes, not loans. We display the answers – in real-time – as green flags plotted on a Google map where every home costs the amount the applicant just told us they feel comfortable paying, indexed annually to inflation, and not a penny more. That sounds radical to a bank with a billion pounds of quarterly profit at their disposal. Meanwhile, charging a fixed monthly rent, related to the amount the applicant feels is affordable, sounds prosaic to three million part-time suburban landlords.
If you can afford to pay £1,200 per calendar month we, the landlord, will let you live in this house over here. If you can afford £1,320 (up 10%) you can live in that house over there. Rent. Location. No deposit. Boom. Done
Our home purchase plans are leases with a negotiated end date, after which the tenant owns outright. Just like a regular rental, the more you feel comfortable contributing monthly, the higher the value of the homes we will finance.
This approach provides the applicant with a straight choice between paying to rent a home they will never own versus paying the same amount to buy a comparable home. Since it takes decades to reach full ownership, a long, patient chat about life goals and long-term spending patterns is the key to tailoring future-proof financing packages.
Starting with the applicant's monthly budget as a fixed assumption, we satisfy our consumer duty at a stroke. Job done. Now, all the applicant need do is choose a home from the list we control. Our pricing is embedded in the algorithm that determines why we plot some homes on the map and not others. If you can think of a way of making pricing more transparent, we're hiring.
Our two-step application form is just the beginning of the getting to know you process. Here is a short list of things banks ignore that renters routinely consider when assessing whether to occupy this house here versus that one over there.
I almost forgot the tricky part. The occupant has to look after the place so, there's a difference between buying a brand spanking new place versus a fixer-upper. That requires Archie to ask a new set of complex questions, such as: would you rather spend your life savings to reduce your monthly contributions, shorten the pathway to ownership, or spruce up the kitchen?
Note that each of these questions requires Archie to know which house you're about to buy – something banks and brokers seem to consider irrelevant. Unless, of course, the value of that house has the wrong ratio to the size of the loan you're contemplating, in which case the banks care inordinately about the choice of house. As it is expensive to value a home and banks (almost) never repossess why do they waste time getting the home valued in the first place? What's the point of pouring time and resources into evaluating redundant collateral unless you're hunting for reasons to say no?
Let's drill down on one simple example where taking time to gather, discuss and parse abundant data adds enormous value to customers
Most aspiring home buyers spend years (not months) relentlessly saving towards a mathematically redundant deposit. If you simply skip the deposit, they can afford to pay more (for decades) to buy a more valuable home by ploughing their recurrent savings into their new home. This kind of innovation doesn't require a single line of code, government support, or taxpayer money.
It's FinTech, Jim. But not as we know it.
Ah, but slow down. With rare exceptions, banks are not allowed to lend more than four and a half times your income. Once you hit the cap, it becomes irrelevant that you're willing to pay more per month – month after month, for decades – than you currently spend on rent. Using this private data (the demonstrable capacity to make regular ISA contributions) profoundly affects the affordability equation. It is not expensive to gather the information and blindingly obvious what to do with it.
Let's see if you can follow the maths.
Jack pays one thousand a month to rent. Without fail, he's put three hundred amonth aside for years. If you make deposits optional, how much can he afford to pay to buy a place, with no change in his monthly outgoings?
Are these esoteric questions or do issues like this crop up in the real world? Well, on average, renters give about a third of their income to their landlords. It's about 40% in London. From a standing start, it can take years to build up a deposit if you put ten percent of your income into an ISA. Ten percent is about 33% of thirty percent. So, if the amount you can finance were linearly related to the amount you feel comfortable paying, you can afford 33% more house with no change in your monthly outgoings and no incremental risk. It takes most people years to save a 10% deposit so this 33% is profoundly important to the home buyer.
Put it this way, if you drop the deposit hurdle, a bank could charge a rate comparable to a rental yield with respect to a home worth 33% more than the value of the home the renter currently occupies. Most renters already occupy homes worth far more than banks are willing to lend, so maybe the product (a loan with an arbitrary income cap) is the barrier to innovation, not the ability to evaluate default risk by parsing income and expenditure patterns.
Our approach to these simple observations is this: design home purchase plans that are not regulated as loans. Boom. Done. Next question
Pushing beyond the loan-to-income cap is a function of service design and data-driven risk evaluation and not a barrier created by the cost of compute or a shortage of chips from Nvidia. Rental yields right now are easily 33% higher than mortgage rates and the word yes is only 33% longer than the word no.
Using open banking technology and data enrichment services to gather spending pattern data in collaboration with specialists like Credit Kudos or Bud Financial was science fiction in 2014 and is very straightforward today. These FinTech specialists can help you analyse the applicant's income and expenditure broken down into incredibly fine categories.
You can pluck out big items, like transport, and then ask the applicant to think about the difference in the cost of their regular commute if they lived in house A versus house B. The route between the applicant's current home and their regular commuting destination is data, right? Figuring out where they want to live requires you to slow down and engage in more yackety yack.
If the applicant is choosing between two potential new homes, their commuting time and cost will change going forward in ways no amount of analysing the past will reveal. So Archie will have to do something technically challenging. Ask the question, then consult Google Maps to evaluate the new choice of transport routes, timing, modalities, and costs.
This gets vastly more complicated if you have joint applicants living in two different locations today with two different regular commuting destinations, with a choice of three modalities (bicycle, bus, pogo stick) which differ depending on which of the three homes they are trying to evaluate as the starting point for their new journey. It takes lots of face-to-face patient conversation to get to the bottom of this transport question alone and some applicants have small kids to look after during this advisory session so maybe a bank branch with a cafe and a creche is not such a strange idea.
Is all this additional yackety-yack worth it, even if it costs nothing at the margin to work with Archie?
We think so. The maths ain't hard. If your home purchase plan is 14,610 days long (that's forty years in the old money) then saving two quid a day allows you to spend £30,000 more for this house than you can for that house because they are located in different places.
The location of these two homes – places you're contemplating but haven't committed to – is data that can only be gleaned in conversation with the applicant. My bet is that Archie can assist a human sales agent in ways that will create a better experience for a buyer (first-time or otherwise) compared to interacting with a chatbot at a distance. Hence, I reckon Archie will reverse the trend of firing human beings. Banks that embrace technology will hire more, not fewer staff. The Taylor Swift of banking will employ a lot of people. She, despite her human entourage, is more profitable than Spotify.
We interrupt this program for a quick word from our sponsors
Quantifying things using daysrent points makes the maths easier for Archie to communicate to his human sales companion as well as to their customers, particularly those with very low financial confidence.
Let's crank through the maths. Pencils at the ready:
If the applicant is comfortable paying £1,200 per month, that's £40 per day. These daily amounts can help people come to grips with super complicated financial strategies, like sacrificing avocado on toast at brunch on a Saturday.
By expressing the answers to this abstruse mathematics as colour-coded pins on a map, Archie can help the young couple evaluate what else they need to forgo to live in that house with the garden close to the park. By pegging the value per point to inflation, small incremental amounts, like £1.70 today hold their value in terms of spending power for decades.
Not everybody is good at maths. Archie is but even he has to dumb it down in order to talk in straightforward terms about a contract where a buyer with no tertiary education is contemplating spending half a million pounds pegged to an unknowable future inflation rate. That's not a typo.
A five percent running yield on a starter home worth £250,000 adds up to half a million of current spending power over forty years. This is not a charitable endeavour.
It's not the low marginal cost of the conversation that will make our approach a better service for consumers; it's the combination of:
In short, it's about taking time to get to know your customer. Getting to know while getting to yes.
And now, another word from our sponsor. I avoid the word 'mortgage' when talking about The Pathway Club because we are not a bank and our home purchase plans, when launched, will not be loans. We haven't formally opened our club membership yet but if you want to get a jump start, go to Canopy.rent and get yourself a RentPassport. Do tell them we sent you.
Banks and other FinTechs affected by these issues should feel free to reach out.
Ike Udechuku | Cofounder | CEO | The Pathway Club