Technology: Direct lending 2.0

Technology-based lending platforms are now increasingly tapping global institutions for capital and serving tens of thousands of small to medium-sized enterprises with loans. Just one well-established platform is now crunching millions of data points to allocate over $100 million per month directly to SMEs on behalf of its capital subscribers.

Meanwhile, fundraising for traditional ‘analogue’ credit funds – larger loans selected entirely by well-educated humans – has plateaued recently. They garnered only $111 billion in 2016 versus $120 billion the year before, according to PDI, chiefly because of concerns about their ability to deploy the funds.

So, because institutional investor demand for credit is set to increase in 2017, the tech-led platforms are now in an ideal position to pick up the baton from the traditional funds. Welcome to ‘direct lending 2.0’, and all the potential for diversity, granularity and high investment yields that it offers.

But how do we ensure that this new and potentially vast capital market is profitable and sustainable as an alternative to the banking system?

Since the financial crisis, over half a trillion dollars globally has been deployed by non-tech direct lending funds into hand-crafted corporate mid-market situations. Some have financed private equity buyouts and some have funded companies looking to grow.

Offices have been opened all over Europe and the US to source and structure face-to-face high-yielding loans without banks’ involvement. Investors have responded, and scores of new asset managers have launched to provide the expertise for pension funds to invest in credit.

However, there are signs of plateauing in what is currently the largest sector of private debt 1.0 – the financing of private equity buyouts. Investors are increasingly concerned about how quickly and effectively the existing cohort of ‘analogue’ direct lending funds can source the handcrafted loans of €10 million to €100 million that is their staple diet. Will private equity M&A activity – still the source of most private debt offerings – remain robust enough and of sufficiently high quality to absorb all the new investor capital?

Most private debt funds in direct lending 1.0 have been providing credit to only 10-20 firms each per year. In fact, the $600 billion of non-bank lending deployed by the traditional non-tech routes in the years since the financial crisis has gone to just a few thousand companies worldwide. This is a tiny fraction of the number of SMEs looking for better credit solutions even when the banking system is functioning well.

It has been an orderly climb to this point for the low-tech, negotiated, large-loan corporate credit sector. The market will continue to function well for the most part. However, if institutional appetite for private credit continues to grow at this pace and if investors are to get the diversity they need, the market needs a refresh.

Tech platforms come of age

Meanwhile, technology is shaking up the processes by which we make decisions and allocate resources in the capital markets. Alongside the growth of private debt funds there has been demonstrable progress for technology platform lending.

Less than $20 billion has been deployed so far a fraction of the total in direct lending 1.0, but this has gone in smaller loans to over 150,000 diverse global SMEs since 2010. Through data-rich platforms, pension funds and insurance companies are allocating directly, without investment banks and private equity firms in between, to much more diverse and granular pools of credit.

One fund may allocate credit to tens of thousands of SMEs in loan sizes of less than 1 million euros or dollars, achieving a yield from the loans at similar or better levels to those in traditional private equity-based direct lending. Moreover, marketplace lenders are showing default and recovery experience in many sectors better than that of the banks.

While the platform lenders’ machines have been learning and improving their own credit selection algorithms, their business models have seen some rethinking, which is healthy. While the models that work will scale up, lead the sector and attract professional capital, others will consolidate and some will fall by the wayside. One would worry if this were not yet the case.

The platforms which attracted equity capital have been able to build their tech capabilities and human resources rapidly. One lender, Funding Circle, has over 500 people worldwide working on credit monitoring, origination, tech, compliance and capital markets functions. This means that it hasn’t actually needed to ‘achieve more with less’, the usual tech company mantra. It is the next phase of the MPLs’ development – providing a channel for institutional investors in large scale – where they will achieve that operating leverage.

The key to that is raising large amounts of pension fund and insurance capital to put to work in now-seasoned lending operations; the best platforms are about to do that, big time.

Pension funds in particular, already familiar with direct lending through the 1.0 providers, are becoming a larger part of the capital allocation to MPLs. This is because for certain types of credit underwriting – namely high-volume, shorter-term SME credit – the powerful new data analytics and communication hubs are bringing about a fundamental shift. Platforms can achieve both tremendous granularity and constant quality of underwriting processes.

The platforms that kept to the SME focus on the lending side and were early to the institutional investor conversation can now show over five years of default and recovery track record. They can also show a good amount of transparency – down to loan-by-loan monitoring data. Typically, the leading platforms are lending to firms that average 10 years of operating history.

All this is music to the ears of institutional investors – they are seeing broad and fast deployment of funds, reduced risk concentration, regular income disbursement and the ability to deploy large funds above €500 million.

But let’s take a step back and carefully revisit our assumptions. Often at this stage in a cycle, when the demand for credit investments exceeds the supply, people start to stretch the core principles of credit investing. The manufacturers of new investment products (typically old products with new acronyms) ride into town. The rest of that story is our living history.

The four most dangerous words in markets are “this time it’s different”. So, while the promise is great, is the technology being deployed by the MPLs really able to shift out the curve that sets volume against quality?

Let’s look at history: what are those core principles in credit markets and how do they get compromised? Then we can assess whether the technology now available is able to give us greater width and scale without compromises.

For this, we must be aware of the four horsemen of the credit apocalypse: inverted telescopes; buying the packaging; maturity transformation; and leverage on leverage.

Don’t invert the telescope

When we invert the investing telescope we zoom out rather than in, thinking we will see useful overall patterns. The problem is, once you fail to see what’s actually going on in each data point in large samples you don’t actually see any patterns, and start to make assumptions – usually over-positive. “US housing prices don’t fall on a nationwide basis” was the most widely-cited fake pattern in the pre-2008 period.

The data-led lenders are able to burrow pretty deep into the SMEs they are lending to and to stay down there. Leading platforms are accessing, analysing and constantly monitoring thousands of data points on each SME borrower. With loan sizes of €50,000 to €1 million, these platforms are typically taking full sight of borrowers’ bank accounts and receiving early signs of deterioration which may impact loan payments.

Some platforms are crunching data reported to the authorities by SMEs that they are not even lending to, to enrich the data learning process. This helps to recalibrate more accurate default prediction models.

This modelling is not the inverted telescope of the subprime market pre-2008. They ran wonderfully complex Gaussian macros on hundreds of thousands of mortgages to help create AAA bond ratings, but down at the coal face where the credit was actually created, they relied on self-certifying salary data and sketchy financial status information.

The SME-focused platforms of direct lending 2.0 on the other hand are using data from the official filings of companies which average around 10 years of operating history to drive their machines’ learning.

Packaging is just packaging

The financial packaging created for investors when mass scale-up of credit investment takes place often exaggerates these fake assumptions. In the subprime bubble of 2005-07, the packaging of new securitisations and CDOs retained the same AAA attachment points, and, if anything, tighter and tighter spreads on their liabilities – supposedly an indication of less risk – while the poison flooded into the system.

So how are the data-led marketplace lenders of today bringing in their investors? In many cases large institutions have been committing to the platforms in remarkably simple investor agreements whereby they commit to capital amounts that are drawn over time and the platform commits to pass through all interest and principal after service fees. The complexity is not in the packaging, it’s in the data – and that is being systematically mined and learned from.

Yes, some of the lending platforms are doing securitisations, which on its own is not a bad thing. However, the mass adoption of securitisation has in the past driven down yields and, in my view, has led to deterioration in underwriting standards just so that the SPV beasts can be fed. In this respect they should be careful, but as long as the credit origination process remains robust, these deals will perform.

Either a lender or a borrower be

Borrowing short to lend long – maturity transformation – is, of course, the perennial bogeyman of credit crises who never actually gets killed off. Marketplace lending platforms, however, are not yet deploying this practice.

Their balance sheets still heavy with venture equity, and if they start to play the maturities – either in their own balance sheets or with SPVs – they enter riskier territory. A virtue of the sharing economy ethos which infuses the platform lender model is that little manipulation of financial structures is done between lender and borrower.

While leverage on leverage may not yet be a feature of the marketplace lenders, there may be lenders on the platforms that are themselves leveraged and borrowing capital to lend to SMEs through the platforms at higher rates. The leading marketplace operators, however, are taking in large pension funds as lenders to the SMEs. These funds’ capital, if anything, is longer-term than the investment vehicles the MPLs are offering.

They are also not leveraged entities. Insurers will, I believe, increasingly allocate to the platforms, where they can get high yields, diversity and granularity. While they may be leveraged organisations, they do not need to lever up their credit investments in any way to meet high return targets.

Check that alignment

A mantra among investors which has proved important in the past is ‘alignment of interest’. It is here that some investors still have concerns about the MPL platforms. When inviting institutions to use the platform to act as direct lenders, the platforms in many cases are not charging the usual annual management and performance fees in the private equity/private debt fund model.

The platform will take a one-off payment each time a loan is originated. A borrower will borrow 100 and receive 97, with the platform taking the difference, while the investor will have provided 100 and is expecting 100-plus annual interest in return. There may also be an annual servicing charge to the investor for the operation of the platform. Beyond these contractual payments, there typically will not be any other direct performance fee charged by the platform to the investor.

So what incentive does the MPL platform actually have to make sure that the loans they make perform well and are managed intensively during their term? Here, some institutional investors more used to the private equity model will be in the mind-set that asset managers only do great if they get rich for it. The lending platforms, however, like banks, above all need to be disciplined. The machines need to source and crunch all the right data. The humans need to act on the data they see. Finding ‘alpha’ and outperforming benchmarks is not the name of the game.

For this reason, the alignment of interest for investors is in the platforms needing to run the most disciplined operations in the market with the lowest default rates for a given return level to be able to attract the most investors. Unlike banks, which levered up equity to increase return for shareholders, the platform’s game is to increase size and therefore profitability through operating leverage – not financial leverage. They only increase size by showing as little volatility in their lending as possible to the lenders they bring into the platform.

Incentives are never perfectly aligned in asset management. For credit investors it seems to me that they have as much of a fit here as they do in credit funds where outperforming a hurdle may encourage extra risk taking.

Mary Shelly wrote that “no man chooses evil because it is evil. He only mistakes it for happiness, for the good he seeks”. As long as we stick to the principles of markets which have been proven indispensable over time, institutional investors can safely ‘seek the good’ of yield, diversity and a regular income through the well-run tech platforms. This should probably be through investing in a combination of the traditional direct lending funds and the marketplace lenders.

Many bellwether institutional investors, such as Railpen in the UK, are already doing so. In the UK in particular the regulators and government itself are firm supporters of the model. I think we will therefore see an investor surge into the marketplace lenders this year as direct lending 2.0 takes shape.

Growth is sustainable ultimately if it is spurred by evolution, not revolution. Financial innovation cannot alter or disregard the fundamentals of credit and of markets. If we can find better ways to apply these principles to new participants we previously couldn’t reach, then we may even help the stagnant economies of the West return to growth.

James Newsome is managing partner of Arbour Partners, a London-based placement agent. This content is sponsored by Arbour Partners.