Why taking the right risks, on the right people, is at the heart of what community lending is supposed to do.
Richard Kirtley, Sep 04, 2026
There is a slightly uncomfortable contradiction at the heart of responsible lending, and it is one I have been thinking about quite a lot over the last few weeks as we have spent more time looking closely at Purple Shoots’ own loan book. On the one hand, organisations like ours are encouraged to reach people who have been excluded from mainstream finance: people with poor or limited credit histories, people receiving benefits, people who live in some of the most deprived communities in the country, people whose businesses are often very new, very small, or not yet trading at all. We are encouraged to lend where banks will not, to back people who do not fit conventional models, and to use judgement where automated systems see only risk. Yet, once we have done exactly that, there is still a tendency to judge whether we are succeeding using many of the same measures that would be used to judge a conventional lender.
Default rates matter. Arrears matter. Portfolio performance matters. Financial sustainability matters. Purple Shoots is a responsible lender and the money we lend needs to come back, because when it does we can lend it again to somebody else. None of that is in dispute. But the more closely I look at the data, the more I find myself returning to a rather different question: what should a successful loan book actually look like when your entire purpose is to lend to people whom the financial system has already decided are too risky?
Since the beginning of 2024, around 470 people have applied to Purple Shoots through our current lending system and around 150 of those applications have ultimately become loans. Those numbers alone say something important about the people reaching us, because the people who come to Purple Shoots are rarely straightforward lending propositions. Many are receiving Universal Credit or other benefits. Many have been unemployed. Many live in communities where access to affordable finance is already severely restricted. Some have poor credit histories, some have very little credit history at all, and many have already been rejected elsewhere before they ever arrive at our door.
These are not unfortunate quirks in the composition of our loan book. They are the reason our loan book exists in the first place.
If our overriding objective were to produce an immaculate set of lending statistics, there are several fairly obvious ways we could do it. We could raise our minimum credit score. We could stop lending to people on benefits. We could require security. We could insist on several years of trading history. We could reduce our exposure to borrowers in poorer communities, or move towards larger loans to established businesses with stronger balance sheets and more conventional evidence of affordability. In all likelihood our arrears would fall and our default rate would improve. We might also look, on paper at least, like a more financially successful lender.
We would, however, have completely failed in our mission.
That is what makes our current loan book so interesting. Despite lending into precisely the kinds of circumstances that conventional finance often regards as too difficult or too risky, around 70% of our borrowers are currently repaying normally or repaying with relatively modest support from our team. If we include borrowers who have only recently fallen into difficulty, many of whom we are actively working with and who may well return to regular repayment, that rises to around 89%. Roughly one in ten loans is in a position of longer-term default or likely write-off.
For a conventional lender, that final number might be the one which immediately attracts attention. For me, the more remarkable number is the much larger one sitting alongside it, because it tells us that the overwhelming majority of people who might have been written off by conventional models as too risky are doing precisely what they said they would do. They are repaying their loans. They are building businesses. They are generating their own income. Some are employing other people. They are spending money within their communities and, in many cases, reducing their reliance on benefits or moving closer towards financial independence.
That does not mean every business succeeds. They, sadly, do not. It does not mean every borrower repays us. They will not. Nor does it mean we should somehow become relaxed about arrears or write-offs, because the sustainability of our model depends upon us understanding those things properly and getting better at managing them. But it does suggest that our understanding of financial risk is often far too simplistic, particularly when it is based almost entirely on what a person has done financially in the past rather than what they might be capable of doing in the future.
There is another kind of risk involved here which is much harder to quantify. The risk of lending somebody £4,000 and not receiving all of it back is visible. It appears on a balance sheet, can be measured in a portfolio report, and can be neatly placed into a spreadsheet. But what about the risk of not lending the £4,000 in the first place?
What happens to the person who has the skills to become self-employed but cannot afford the equipment they need to begin trading? What happens to the parent who desperately needs flexible work because conventional working hours simply do not fit around their family? What happens to somebody leaving unemployment who has a credible route into self-employment but does not have several thousand pounds sitting in a savings account? What happens to the entrepreneur whose only alternative is to use extremely expensive credit because no affordable lender is willing to take a chance on them? What happens to the person who, after another rejection from the financial system, simply gives up?
Those outcomes have costs too. They have costs for individuals and families, but they also have costs for communities and for the taxpayer. They just happen to be much more difficult to capture than a missed loan repayment.
We have had a particularly tangible reminder of this over recent months through the Evergreen funding originally provided to us via BCRS in the West Midlands. Purple Shoots received £41,000 to invest into small businesses in the region and we have now deployed all of it. In the grand scheme of business finance, £41,000 is not a huge sum of money. There are funding announcements made every week which involve tens or hundreds of millions of pounds. But what I find so compelling about microfinance is that the scale looks completely different depending upon where you are standing.
Viewed from Whitehall or the City, a £3,000 or £4,000 loan is almost invisible. Viewed from the kitchen table of somebody who needs a piece of equipment, a vehicle, a deposit on premises, stock, insurance or simply enough working capital to get through the first few months of trading, it can be genuinely transformational. The £41,000 we have deployed has reached real people attempting to build real businesses. It is helping people generate income, spend money locally and create opportunities for themselves which might otherwise not have existed. Better still, because the funding is evergreen, when those loans are repaid the money can be lent again, and then again after that.
This is why I think the way we talk about the success of community finance needs to become more sophisticated. There is significant momentum behind CDFIs and community lending at the moment, and that is undoubtedly welcome. There is growing recognition that conventional finance does not reach everybody, and that there is a role for lenders prepared to work differently. But as the sector grows, there is a danger that we create incentives which gradually encourage community lenders to behave more and more like the institutions they were originally created to provide an alternative to.
Scale is important, but scale and impact are not necessarily the same thing. A lender might be able to point to £50 million of capital deployed and that is unquestionably significant, but I also want to know who received it. I want to know where they lived, what their financial position was before receiving the loan, whether another lender had already declined them and whether they would realistically have been able to access finance without a CDFI. I want to know whether their income increased afterwards, whether their business survived and whether it created employment for anyone else.
Of course we should measure defaults. We should measure arrears and repayment performance, and lenders should absolutely be accountable for the capital entrusted to them. But we should also measure additionality. We should measure how many borrowers had already been rejected elsewhere, how many were unemployed before starting their business, how many were receiving benefits, and how many live in the most deprived parts of the country. We should measure businesses created, jobs sustained and household income generated. We should look at the economic activity that results from every pound of capital deployed.
Perhaps we should also start asking a more uncomfortable question: how much genuine risk was a lender prepared to take in pursuit of its social mission?
One of the things our own data continues to teach us is that people are much more complicated than a credit score. Someone can have poor credit and still build a successful business. Someone receiving Universal Credit can become an excellent borrower. Someone living in one of Britain’s poorest neighbourhoods can create jobs and contribute substantially to their local economy. And someone whom an algorithm might reject in seconds can look very different after an hour-long conversation in which you understand what has happened to them, what they are trying to achieve and whether their plans really stack up.
None of that means ignoring data. In fact, Purple Shoots is probably using and analysing more data now than at any point in its history, because we want to become better lenders and better understand what genuinely predicts repayment and what does not. But the purpose of that data should be to improve human judgement rather than eliminate it.
There will always be people we back who do not succeed. There will always be businesses which fail and loans which are not fully repaid. If there were not, I think I would begin to wonder whether we were taking enough risk at all.
Our purpose is not to build Britain’s safest loan book. Our purpose is to give people who have been excluded from opportunity a fair chance to create one for themselves.
That requires risk. Not reckless risk, and certainly not irresponsible lending, but thoughtful, informed and deeply human risk. It requires accepting that somebody’s past financial circumstances do not necessarily define their future, and that sometimes the job of a responsible lender is not simply to identify the people most likely to repay, but to work out which people deserve the chance to prove that they can.
Behind every percentage point on a default report is a person. But there is also a person behind every application we decide is too risky. Perhaps, then, the success of a community lender should not be judged simply by how much money it manages to protect.
Perhaps it should also be judged by how much human potential it is prepared to unlock.