Unit Economics of Lending & how Tech can improve them (Part 2/3)

Unit Economics in Lending: In part 2 out of 3, we'll look into Cost of Servicing and Cost of Financing. How can tech help you improve on each of these metrics? What are other companies doing?

Philip Kornmann
Philip Kornmann

This is a series of 3 parts in which we go through the Unit Economics of Lending and how tech can help you improve on those metrics. In this part, we’ll dive into the Cost of Servicing and how innovative payment methods can help you not only improve on that metric, but also on others. We also take a closer look at the Cost of Financing metric. In Part 1, we focused on the Cost of Acquisition and in our next post we’ll focus on Defaults, Expected losses and your Margin.

(For those of you who read the first part of the series, you can skip through this first part and go directly to Average Cost of Servicing as this will be a bit of repetition to you)

Unit economics in Lending

Unit Economics is a common way to break down any business into smaller components in order to see how it is performing. If you have ever heard the terms Cost of Acquisition and Customer Lifetime Value, you have encountered the concept before. 

In the following section we will break lending down into its Unit Economics. When doing so, we will focus on the average figures, i.e. the economics per unit as the name entails. This can be done in multiple ways, the below version is our way of doing it (if you think we have missed something, send us an email and we are happy to update the post). 

We will use three examples, based on approximate figures that we have observed in live examples of lending businesses. For each metric, we will give some examples on how tech can help to improve the specific metric.

In the above example, we have assumed that each loan runs for a 12 month. As you can see, the metrics differ quite a lot between the three examples. That is not to say that one is better than the other on all metrics necessarily. It can be a strategy to be the high-cost lender that is approving a larger number of clients than what a low-cost lender can do. Below, we dive a bit deeper into these metrics and how tech can help you improve on each one of them. 

Average cost of servicing

The servicing of a loan refers to the administrative costs associated from the time when the loan was issued until it was fully repaid. This mainly includes invoicing, collecting monthly payments, maintaining records of balances, sending reminders etc. The servicing costs does not tend to be a large cost for lenders. However, as you can imagine quite a lot of these tasks can be automated through technical features as they are pretty standardized in their nature. Also, there are different operational tactics applied here, which affects the user experience, can have an effect on your Expected loss and decrease the administrative work associated with servicing. 

(Re)Payment method

The payment method that you offer your clients to use when repaying the loan can have an effect on the user experience and the retention rate as mentioned in Part 1 of this series. For example, the invoice you send to your client can include additional information and a link to a user-panel, which engages the client and allows for upselling (i.e. offering the client a new loan or a top-up).

Another important effect from payment methods is that it can have an effect on your Expected Losses too. Making the act of paying an installment on a loan as easy as possible can have effects on your Expected Losses. Below are a couple of interesting examples, and how tech can help you to implement them:

Direct debit:

Froda, a Swedish lender, is using direct debit to charge interest and installments from their clients. Each day, a small amount of the loan is repaid along with the cost of the loan. Direct debit is per se nothing new and is a common way to secure routine payments, but applying it with daily payments for loans is quite uncommon. From Froda’s perspective (we are guessing here) it allows for securing payments quickly and early indications on payment arrears. The clients on the other hand, might like the fact that they don’t have to spend time on paying invoice, but can also dislike the additional accounting work it requires.

Tink's direct debit solution using PSD2

Alright, so how can tech help you with Direct debit? One innovative solution is to use service providers acting under the PSD2 framework to access your client’s bank account details, like Tink for instance. This allows you to get the necessary information to start charging the client using direct debit, as opposed to the alternative of having your client finding the bank details and filling them in manually.

Invoices with payment option in them

Sending invoices by email is nothing innovative. However, to make the repayment easier for the client one can include payment options directly in the invoice. E.g. sending the invoice as a link in an email, which takes the user to a page with the invoice and a button that launches a payment solution like Trustly, allowing for direct payment from the client’s bank account. 

Bexio's invoices sent to clients with various payment options

System - payment reconciliation, reminders, and recalculations

Scaling up as a lender means that you need to keep track of hundreds or thousands of balances and payments each month. In order to do so, it is important to have a functioning system in place that allows you to reconcile what loan installments that have been paid or not, automatically remind those customers that have not paid, and recalculate the loan balance and payments to be made for this group of customers.

Luckily, there are software providers, such as Mambu, that offer systems that allow lenders to do all of the above without having to program each functionality from the ground up. However, this clearly comes with a fee and does require a pretty complex implementation to make the user experience good and to have efficient operations. We have seen good examples of implementations (N26 implementing Mambu) and bad ones (no, we’re not gonna mention these).

To summarize, the cost of servicing does not tend to be a major component in the unit economics of an online lender. As you can see in our examples above, whilst the difference is large in relative terms, Lender the absolute difference between Lender 1 and 3 is only 900, which does not have a big impact on the overall result. However, the tactics you apply for your servicing will have an impact on other areas of your business as it involves payments, reminders and how you handle clients that are not paying their loan installments. Areas which can have a large impact on your overall result. 

Average cost of financing

Common sources of financing

The cost of financing is what a Lender pays for the funds that are lent out to the borrowers. Until a lender has proven it’s credit management skills this tends to be the shareholders’ money, i.e. equity. However, as this is expensive (equity holders expect a higher return than what debt providers tend to) and puts a limit to the portfolio size of the lender, it makes sense to look for cheaper sources of financing. Below are some common options that we have observed amongst online lenders we have worked with

  • Loans from the lenders’s shareholders
  • Loans from financial investors
  • Credit facilities with a financial institutions
  • Bonds
  • Off-balance sheet solutions (securitization using an SPV for instance)
  • Consumer deposits

A couple of examples from Sweden

Below are a couple of examples that we have observed on the Swedish business loan market (these facts are public, so no, we are not revealing any information that we have received from clients and partners):

Qred - A plethora of sources

Fintech player Qred started their lending business in 2015 by using their equity and money that they lent from their shareholders. After having operated for 1-2 years, the company was able to set up a credit facility with a financial institution. Our guess is that the financial institution they partnered with required a minimum level of equity being used in the lending operations (read skin in the game), why they also raised equity financing from their shareholders. To avoid being too diluted, the company started to offer Preference Shares to the public, without any voting power but with an interest payment of 7% and the possibility to be resold to the company at any time. 

Qred's preference shares, without voting rights but with an interest payment of 7%

Later on the company issued a 100 MEUR bond to financial institutions priced at EURIBOR + 8.5%. In 2020, the company has started to migrate the majority of their portfolio to an off balance sheet solution. Though we don’t know the cost of financing for all the sources, we think it is a rather interesting example of how it can vary over time for fast growing and innovative players.

Froda - accessing the sweet consumer deposits

Froda, another Fintech player in the business loan space, received an equity investment from the Swedish bank Svea pretty early on. Our guess is that a credit facility was set up with the very same bank along with that equity investment (though this is pure speculation) to fund the lending activities. 

Having worked diligently over a few years, making use of bank account data being accessed under the PSD2 framework, Froda had grown enough to apply for a license as a financial institution. In 2020 the company was awarded the license and as a result could offer consumers deposit accounts that are covered by the governmental bank guarantee. In principle, it means that consumers can set up a savings account at Froda and deposit up to 100 000 EUR risk free. For many lenders, this is the holy grail of financing sources as consumers tend to accept very low interest rates on their deposits (Froda offers a mighty 0.85% on their savings account…).

So how can tech help here?

Accessing cheaper sources of financing is done by building a larger loan portfolio that yields a good return over time and use as a proof of your abilities when raising new financing as well as in applications to the financial authorities for a more advanced regulatory license. In other words, you need to be good at 1) growing your loan portfolio (read sales and marketing), 2) predict the default rate, 3) selling the portfolio performance to financing providers, and 4) managing your financial regulations work (AML/KYC) in the case you aim to accept consumer deposits.

Being good at selling your portfolio performance to financing providers is a topic of discussion for a different blog post. Managing your KYC/AML operations is also something for another post (you can read more about here). Alright, so what can tech really help with here? It can help you to grow your portfolio, a topic we discussed in the first part of this series under the Cost of Acquisition section. Tech can also help you predict the default rate, which we will dive deeper into in the third and last part of this series. 

Applying it to our examples

Taking a quick look at the table at the beginning of this post we can see that the difference in the Cost of Financing for the three examples is up to 14%. Having a Cost of Financing of 15% is not uncommon for lenders targeting high risk segments or lenders who have just started to operate and don’t have a lot of data to be used to prove their default prediction abilities. However, as we have discussed above, it is possible to decrease this metric a lot by working your way towards cheaper sources of financing. 

Getting low cost financing is a must if you are targeting a value proposition with a low interest rate to your clients. However, the market for loans with a bit higher risk is also a market to be served and does not require as low costs when it comes to financing.

That's it for Part 2 in this series. In the next part, we'll look into Defaults, Expected Losses, and Margin. Stay tuned! 

Published by
Philip Kornmann
August 2022