Honeymoon is over?


Since the first days of operations FinBee and its investors lived like a lovers in a honeymoon. Now, as interest rates dropped significantly, some of the investors started to grumble. So, is the honeymoon over?

Darius Noreika, COO of FinBee conducted a portfolio analysis to better understand if there are objective reasons to be dissatisfied with results one can achieve while investing.

Please read, discuss and share.

Analysis of the Loan Portfolio of FinBee

Rating model

Due to decreasing interest rates at FinBee, I have decided to perform an analysis of loans portfolio of FinBee. Loan portfolio of FinBee is pretty “young” and as you all know bad debts usually peak after some time after loan origination. For my analysis I have used loans that were originated from August 2015 to March 2016 only and loan statuses of those loans at the end of December 2016. All included loans were originated 10-17 months before. The average term of loan at FinBee is 32 months, so loans used for my analysis have already reached half of their term.

FinBee granted 467 loans in the first 8 months of operations. You can see the structure of loan portfolio according to risk category in the table below.


The common practice for loan portfolio analysis is to show all data in proportions rather than absolute figures. The table below you can see the same data in proportions:


Chart of late loans according to risk category:


As you can see in the chart above, lower rating categories has way more late loans that higher rating categories (A* is the highest rating category, D is the lowest rating category). This chart is very good example that rating model of FinBee is quite accurate and assigns loans to appropriate risk categories properly.

For loans originated from August 2015 to March 2016 investors earned interest of 21% after loan losses. Such interest rate seems to high according to the risk investors are taking. Currently lender interest rates are lower than it used to be. Taking into account average interest rate of loans originated in December 2016, interest rate after potential loan losses is still 14%. It must be noted that I haven’t taken Compensation fund into account in my analysis. Compensation fund reimburses all more than 61 days late monthly instalments. This increases investors’ return, but as Compensation fund is no an obligation of FinBee to cover all loan losses in the future, it doesn’t not eliminate credit risk.

Screen Shot 2017-01-23 at 16.17.20.png

In general, the quality of the loan portfolio of FinBee is good, the share of defaulted borrowers is small compared to lender interest rates.

Evaluating the quality of the loan portfolio, I was worried that the quality of new granted loans might be decreasing due to increasing lending volumes. The best way to find out whether my worries are true or not is to compare monthly data of loans with late payments after the same time from loan origination. You can see this data in the table below. The table includes the share of loans with late payments after 3, 6, 9 and 12 months after loan origination(Click on the table to zoom):


The table is quite complicated to read, so I decided to show the same information in the chart:

Screen Shot 2017-01-23 at 16.21.27.png

As you can see trends of the chart shows that the quality of new loans granted isn’t worsening. Due to low number of loans there were quite heavy fluctuations in the first months. In later months the quality of new granted loans was improving (share of loans late for at least 3 business days was decreasing).

The term of the loan, the loan amount and DTI

The interest rates usually drop significantly below target interest rate for loans with short term and small amount. I wanted to find out whether this is caused by emotions or economic reasoning, so I decided to split loan portfolio into categories by loan term and loan amount.

Number of loans with late payments according term of the loan:

Screen Shot 2017-01-23 at 16.27.31.png

Share of loans with late payment according term of the loan:

screen-shot-2017-01-23-at-16-23-25 Visual chart:

screen-shot-2017-01-23-at-16-28-37It is obvious that loans with longer term are late more often than loans with a shorter term.

Number of loans with late payments according loan amount:


Share of loans with late payment according loan amount:


Visual chart:


* It must be taken into account that number of loans with such loan amount is very small, so any significant conclusions shouldn’t be done based on this data.

Analyzing loan portfolio according to loan amount, the situation is similar to the situation according to loan term. The smaller loan amount is the less loans with late payment there are. This is not surprising at all, because it is more difficult to pay bigger monthly instalments for bigger loans on time than smaller ones.

Analysis of the quality of loan portfolio according to loan term and loan amount shows that loans with shorter terms and smaller amounts are less risky than loans with longer terms and bigger loan amounts. So, I can make a conclusion that there is an economic reasoning behind lower investors’ interest rates for small loans with short loan terms.

The most surprising results came from analyzing the quality of the loan portfolio according to DTI (Debt-to-income ratio). DTI is usually considered as one of the main indicators that shows borrower’s ability to pay his/her monthly instalments without difficulties. My analysis shown that DTI doesn’t have significant impact on meeting your financial obligations on time.

Number of loans with late payments according to DTI:


Share of loans with late payments according to DTI:


Visual chart:


My personal expectations was that DTI has an impact on loans with late payments, this chart clearly shows that my hypothesis was wrong.

The appetite of investors

I wanted find an answer whether lending below target rate increases credit risk taken by investors. I decided to look at difference between target rate and lender interest rate and how this difference affects the sale of late loans. Results were really surprising. The bigger the difference between target rate and lender interest rate is, the less borrowers are late to pay monthly instalments on time.

Number of late loans according to the difference between target and lender interest rate (negative difference means that investors are lending with a lower interest rate than a target rate):


Share of late loans according to the difference between target and lender interest rate:


Visual chart:Screen Shot 2017-01-23 at 16.43.09.png

FinBee refinances those borrowers that have financial obligations (especially for fast credit providers with high interest rates), but they are still able to meet their financial obligations on time. Since FinBee offers better loan conditions for its borrowers (interest rates are lower compared to fast credit providers), it is easier for the borrowers to pay smaller monthly instalments for FinBee. So it seems that when investors are willing to lend for lower interest rate, they reduce overall risk level of the portfolio.


The main findings from my analysis were:

  1. Rating model of FinBee is working perfectly fine.
  2. The quality of the loan portfolio is improving while loan origination volumes are growing.
  3. Investors behaviour to lend with interest rate below target rate is reasonable and justified by economic logic.
  4. DTI isn`t the main indicator that shows the borrower’s ability to meet his\her financial obligations on time.


2 mintys apie “Honeymoon is over?

  1. Thanks for the data and your analysis.

    You say that the interest rate that investors have earned (21%) seems too high for the level of risk assumed (this is very subjective), but surely it is also about competition with other lenders. If borrowers in Lithuania can borrow on FinBee platform cheaper than from other lenders, then they will be happy and FinBee investors will be happy because the rates are higher than they can earn on other platforms. However, with rates falling to 14%, lenders can earn this on other platforms with less work (not having to monitor auctions and re-bid), so the interest rate at this level is less appealing.
    I also note that the interest rate after losses in December on C and D loans is less than that for A and B loans. This is the reverse of what I would expect as it would be expected to receive a higher return for taking more risk. This implies that the interest rates for C and D are too low.

    I agree that some investors don’t want to invest in the longer term loans, but I think that the reason behind lower interest rates for smaller loans is simply the mechanism of the reverse auction. With lenders bidding the same amount irrespective of the loan size, the interest rate falls faster if the loan size is smaller.

    You say you were surprised to see that fewer borrowers were late with repayments if the lender interest rate was lower than the target rate. I don’t think that this is surprising (smaller repayments are less arduous), but investors do not want to loan money for free (in the limit) to decrease the risk of the borrowers not repaying! Paying higher interest rates may result in more defaults, but the overall return to the investor may be higher (up to an optimum level of interest rate).


Parašykite komentarą

Įveskite savo duomenis žemiau arba prisijunkite per socialinį tinklą:

WordPress.com Logo

Jūs komentuojate naudodamiesi savo WordPress.com paskyra. Atsijungti /  Pakeisti )

Google photo

Jūs komentuojate naudodamiesi savo Google paskyra. Atsijungti /  Pakeisti )

Twitter picture

Jūs komentuojate naudodamiesi savo Twitter paskyra. Atsijungti /  Pakeisti )

Facebook photo

Jūs komentuojate naudodamiesi savo Facebook paskyra. Atsijungti /  Pakeisti )

Connecting to %s