Lessons Learned & Improvements To 4thWay Research & Ratings

I'm super proud of what 4thWay has done for many lenders and of our record. We have made some mistakes, but our ratings, research and guidance has been very accurate. More accurate than it needs to be for lenders to comfortably make satisfactory returns over the past 11 years.

And the ultimate mistake of causing an overall loss on diversified lending portfolios is still a long, long way off.

Nevertheless, it bothers me a great deal when I let people down and that has happened recently. Here's what happened, why it happened, and what steps we have already taken for 4thWay to do even better in future. I'll try to be brief.

What was the error?

As I was leading the research into CrowdProperty, I took too long to spot that the conditions were worsening to the point where lenders might now potentially lose money overall in one or two year's worth of loans.

On average, it's possibly going to be as much as 6% in losses after interest earned in one of those years. That's not an annualised figure, which would be worse, but rather an estimate of the potential total losses to come on the set of loans approved that year.

Since no-one is “average”, individual lenders' losses will be greater (or lesser), depending on whether you lent across lots of different loans and what particular set of loans you had.

If that does indeed end up being the final result, it will basically wipe out a year's gains made in one of your other lending accounts, assuming you're sensibly, evenly spread across a basket of them. (Read the latest CrowdProperty Review for more.)

Historically, 4thWay's assessment of what was likely to become an actual default has been a lot better than CrowdProperty's own. For example, three years ago CrowdProperty didn't log any loans as being troublesome, whereas we were estimating 7% as de facto bad debts, which turned out to be much closer to the truth for the loans issued up to that point.

However, as CrowdProperty hadn't acknowledged those problems to us, but were otherwise very open to us, I failed to pull the threads together to see the signs that this was just the beginning. I'm sorry about that.

To be clear, CrowdProperty's record is taking a substantial knock, but it's still an ongoing lending platform. It's not collapsed and 4thWay has always succeeded in giving at least a year's notice to lenders before any peer-to-peer lending provider that we had been tracking went under in any serious form of disgrace. As we did with the likes of Wellesley, Lendy and FundingSecure.

Great. But that doesn't stop me feeling intensely motivated to do even better in future. So it's important to consider what went wrong.

Why the error occurred

Reduced resources

The biggest reason is that roughly 18 months ago I started shrinking 4thWay with a view to closure and so the resources for assessments and analysis became too stretched.

This decision has been reversed, as I found a way to make 4thWay viable, but the damage was done in terms of falling behind on our coverage and due diligence.

An untapped resource

We also weren't getting all the insights we should have done. Specifically, 4thWay has been so focused on bank-like and investment-fund-like assessment techniques that we were not using another useful resource available in this space: the crowd.

If we had a system for connecting with more individual lenders, having our ears to the ground, it's likely that we would have spotted the gravity of the intensifying issues far sooner.

Background noise filtering

CrowdProperty's borrowers had unquestionably been hit by an absolute tsunami of bad news that was out of their control. It's no surprise that they are suffering and it impacts lending results.

However, in hindsight, we have now identified signals buried in the quantitative and qualitative data and information we receive that we had missed. It could have aided us to understand the other factors going on specific to CrowdProperty.

Delayed reassessment

It took far too long to reassess CrowdProperty again.

A large part of that is simply that it took longer for CrowdProperty to give us the information we asked for and to get access to its key people. There's little that can be done about that.

And partly it was just a very big reassessment, and that takes time.

But unquestionably I could have ensured that 4thWay's first steps in the reassessment started faster and with more energy.

What we're doing now

Doubling resources

The decision to close 4thWay has now been fully and indefinitely reversed as of spring this year, as has the shrinkage of resources.

Indeed, I have doubled the resources for ratings and research, so that it is now more than it has ever been. I have also shifted our focus so that more of the research time is dedicated to monitoring, assessing and reassessing platforms and less is on other kinds of research into the wider field.

These changes will help us catch up with the few remaining providers that we're still behind on soon and stay ahead thereafter.

It also means more time and depth going into more of the research we do behind the scenes.

Incorporating lessons learned into our internal analysis

We have added to our rules-based approach in reassessing providers. All of the signals that we have identified that we had missed are now incorporated into a standardised checks. For example, when there are subtle changes to the characteristics of loans being approved.

We've started automating some of those tests with our own-made, internal assessment software.

More requests for answers where necessary

For P2P lending providers that are relatively slow to recognise bad debts, we are scheduling more regular requests for qualitative information and/or documentation to better keep an eye on what's happening.

Later, I'll review how responsive platforms are to these requests, because it can often be time-consuming for them to provide what we need. I'll consider how effective it has been, whether what we are getting is sufficient or what other measures need to take place.

4thWay PLUS Ratings model upgraded

We have also adapted our 4thWay PLUS Ratings model to incorporate some of the rules-based changes directly into it, so that it better reflects changing inherent risks.

With help from the crowd

All 4thWay's contact with lenders in the past has been ad-hoc. Now, I personally talk directly with individual lenders and I'm working on building regular contact.

These volunteers allow me to see and feed back to the team what's happening through the eyes of the people actually putting their money on the line.

It's possible that separating valuable insights from lenders' individual histories and anecdotes, and all the rest of the background noise, will be a messy affair. And in this form of research I am starting more or less from scratch, so I'll need to practice it and measure the results. But it's got started.

If it ultimately improves further on our past record, this will be a success and I have high hopes that it will be worth the resources put into it.

How much difference would those measures have made?

Had the above measures already been in place, I believe that we would have at least been able to warn you more strongly of the building issues specific to CrowdProperty early enough to make a difference.

I'm not convinced that we'd have been able to forecast soon enough that the situation would deteriorate as badly as it did. But the increased warnings that we would have been able to provide could well have educated you enough about the changing risks at CrowdProperty. That might have enabled you to reduce your lending punctually enough in those years to reduce your risks.

I continue to put you lenders above everything else and I really want to be responsible for helping you do well and feel well with your lending and your results. I'll keep trying to do my best!

Pages linked to above

CrowdProperty Review.