Understanding how Zopa risk management works
Zopa screens prospective Borrowers and, based on risk analysis, advances money on behalf of Lenders to those approved by the process.
The assessment assigns approved Borrowers to a Market (A*, A, B, C or Y) or rejects the Borrower as being too risky. Market assignment determines the interest rate a Borrower will receive based on the minimum offered by each Lender to that Market.
Once a loan has been made:
- Zopa monitors the individual Borrowers for the earliest signs of possible Default on the loan and subsequent bad debt. This is normally indicated when a Borrower does not make a payment on time (i.e. a Borrower falls into Arrears).
- When a Borrower misses a payment, Zopa begins account management activity to bring the account back up-to-date as soon as possible. This process is effective in bringing many customers back up-to-date.
- If Arrears persist Zopa begins collections activity using the expertise of a collections agency, looking to get the Borrower to pay as much as possible. However, at this stage Zopa expects that a proportion of the outstanding capital and interest will not be recovered.
- If a borrower in Arrears does not make or honour an alternative arrangement to repay the Arrears, Zopa will issue a Notice of Default. The borrower has 14 days to respond to this letter, during which no further collections activity can take place.
- At the end of this 14 day period, if no repayment of the Arrears or formal arrangement to repay has been made, the borrower is said to have fallen in to Default.
- Once a borrower falls in to Default, collections activities do continue, but the probability of collections is relatively low.
Zopa tracks the performance of each Market (A*36, A*60, A36, A60, B36, B60, C36, C60, Y36 and Y60). This includes observing the markets' payment performance for each month's business over time. From this Zopa predicts the lifetime bad debts of a tranche of loans.
Managing expected Arrears & Defaults
Initially, Zopa's Borrower approval criteria were oriented towards virtually no Arrears or Defaults. Over time, Zopa has modified the approval criteria to allow a more active market that includes the expectation of Arrears and Defaults, while continuing to provide a range of acceptable net returns to Lenders. Even with this change, Zopa has historically had a low level of Arrears and Defaults.
Zopa therefore uses a small amount of internal data in combination with expert knowledge from the industry to develop expectations of how Defaults will develop in every Market. The following chart shows a Typical Bad Debt curve illustrating how Arrears and Defaults mature over the life of a tranche of loans.
This shows:
- Arrears grow fastest over the earlier months of a loan portfolio.
- Defaults and Arrears continue to grow over subsequent months, but more slowly to a point where they plateau.
- Some of the Arrears result in Defaults, whilst other Arrears are recovered.
- The amount of Arrears experienced reduces considerably as the portfolio matures since more of the loan capital has been repaid.
- Borrowers who have paid up to this point are very likely to continue to pay each month.
How Zopa derives the Expected Bad Debt curve for a market
Zopa continuously monitors the actual performance in each market (shown in detail here). Using a combination of detailed actual performance (which has been limited to date), the known general shape of the Typical Bad Debt Curve and expert knowledge in lending markets, Zopa projects a specific Expected Bad Debt Curve for each market.
From time to time, as market performance and market conditions change, the Expected Bad Debt Curves are modified accordingly.
Providing bad debt information to Lenders
Data correct as at 12th March 2010.
Zopa uses the Expected Bad Debt Curves in combination with actual market performance to provide Lenders with information that allows them to compare actual performance against expected performance. The following chart shows how the loans that have been issued in all our markets (excluding listings) have performed, compared with our expectations.
The bars represent:
- Grey: Expected performance (sum of Arrears and Defaults)
- Red: Actual Defaults
- Orange: Expected Defaults from current Arrears (it is assumed 50% of Arrears will be recovered)
The bars are arranged in to the following tranches:
- Last 12 Months: The weighted average performance of all the loans made to 36 and 60 month markets (A*, A, B, C and Y) in the last 12 months.
- From 1 to 2 years ago: The weighted average performance of all the loans made to 36 and 60 month markets from 13 months ago until 24 months ago.
- From 2 to 3 years ago: The weighted average performance of all the loans made to 36 and 60 month markets from 25 months ago until 36 months ago.
- From 3 to 4 years ago: The weighted average performance of all the loans made to 36 and 60 month markets from 37 months ago until 48 months ago.
- The height of the grey bars reflects the Expected Bad Debt Curve. Imagine twelve of these each representing a month in a year. The "Last 12 Months" tranche is 12 curves combined giving the average expected performance of loans issued in the most recent 12 months. By the end of the first year, lenders should expect to see a significant proportion of the ultimate bad debt to have materialised (mostly as Arrears at this point).
- The actual experience shows higher levels of Arrears and Defaults amongst loans disbursed from 1 to 2 years ago when compared to other tranches, and performance has been brought back within expectations on loans issued over the last 12 months. This is a result of a combination of the worsening national economic conditions and a deliberate adjustment by Zopa to produce bad debts more in line with predictions.
Impact of bad debt on Lender returns
Lenders' actual net returns are determined by two factors:
- Annual Interest Rate at which the loans were offered.
- Annual Interest Rate Impact of the ultimate bad debts.
Zopa uses the Expected Bad Debt Curves to project the ultimate bad debt for each Market. Zopa then works out what this means in terms of the impact on the annual interest rate charged by a Lender; in other words, how much interest would just cover the bad debt. Alternatively, it can be looked at as the premium that Lenders should add to their required annual interest rate to deliver themselves a satisfactory net return (after both bad debt and fees).
As an example, ultimate bad debt of 2% yields the following Annual Interest Rate Impact:
| Term of the loan | Ultimate bad debt | Annual interest rate impact |
|---|---|---|
| 36 months | 2.00% | 1.289% |
| 60 months | 2.00% | 0.782% |
So, if Lenders lent in a 60 month market with:
| Annual Interest Rate | 8.000% |
| Ultimate Expected Bad debt | 2.000% |
| Annual Interest Rate Impact (see table above) | 0.782% |
| Annual Lender Fee | 1.000% |
| Net Annual Return | 6.218% |
At the individual Lender level, this net annual return can vary widely. Looking at a simple example (not real data):
- £1,500,000 was lent in the B60 market in that month.
- There are 300 loans of £5,000.
- 6 loans went to Default (the expected 2.0%, assuming no repayments at all in this simple example).
- 2 Lenders lent the same amount to 100 Borrowers in this Market.
- Then, it is possible that:
- One Lender may find that all 6 Borrowers that Defaulted were among his 100. His bad debt rate would be 6.0%, with an Annual Interest Rate Impact of 2.317%.
- The other Lender finds that none of the 6 Borrowers that Defaulted are in his 100. He would have experienced no bad debts, with no Annual Interest Rate Impact.
Providing Lenders with tools to manage net returns on new loans
Zopa uses these Annual Interest Rate Impacts to calibrate the Customized Lending Page on the web-site, guiding Lenders on how to estimate their expected returns after Defaults and fees.
Zopa modifies these expected bad debt estimates on the Customized Lending Page in January, April, July and October if actual experience and analysis suggests that the change would better reflect expected performance at that time.
Changing approval criteria over time
Zopa uses the Expected Bad Debt Curves and the detailed data behind them to guide the approval criteria for new Borrowers in each market.
Given market changes and Zopa's growing experience, changes to approval criteria are made from time to time to ensure that the portfolio performance will reflect the expectations of Lenders while still offering a competitive product for Borrowers in the market place.
So, for example, Zopa recently tightened the approval criteria for C markets, based on market conditions and the performance of C Market Borrowers over time.
Click here to view the current annual bad debt estimates.