Impact of the new Bondora Rating system

As of last night, Bondora took the first step to push the new credit system live. As of now you can see the old scoring (ABC600-100) alongside the new scoring system (AA A B C D E F HR) on the market. Leaving aside the fact that the update broke a significant part of the functions accidentally and removed some on purpose, how does the new scoring system impact investors?


Tauri’s visualization of the link between credit groups and risk

Automatic vs manual investing

For those investors who have always invested automatically using portfolio managers very little will change. If you ran a very conservative strategy, which is likely if you have a small portfolio, you can just accept any general portfolio manager setup that Bondora offers you and you’ll be set. Investing will be just as passive as before, and you’ll end up being safer from risk than you are now.

For those like me who do a part of their investing manually and run portfolio managers that are a bit more high risk than the default ones, more is going to change. Firstly, the returns for active investors will likely drop because more market inefficiencies will be eliminated. Secondly, investing will likely become more passive due to less need to scan through data provided because the new system works on partially hidden data, which means that any analysis you are running will be insufficient. Thirdly, there is definitely a lot of potential that once the new scoring is finished, that our portfolios might prove to have some very dangerous combinations of loans, such as a huge portion of HR loans.

Different process of analysis

Due to a lot of data being hidden, then it will be up to the investors to trust Bondora to rate the loans well. For example, if you look at the top 3 Estonian loans listed by potential returns onthe market right now, this is what you see:

12.73 return – HR (A600 unverified – 38%)

9.87 return – B (A1000 verified – 15%)

9.3 return – C (A1000 verified – 15%)

Now, if you look at the 3 worst Estonian loans listed by potential returns on the market right now you see something equally confusing:

0.43 return – HR (A700 income verified – 34%)

1.44 return – HR (A600 unverified – 38%)

3.56 return – D (A1000 verified – 15%)

When looking at such comparisons, it’s clear to see why you need to be able to trust the sytem Bondora made. Just looking at the potential interest or credit group is not enough to be able to predict the risk group a loan falls under. Looking at an ordinary A1000 loan, the different risk factors that Bondora takes into account can make it fall under B, C or D rating while the interest and income is the same, but potential returns for those examples differ by a magnitude of 3 times. Same for the two A600 loans – they are both the top and bottom performers listed on the market currently.

Long term impact

– Social lending will likely become more more passive (at least for me personally)

Returns will even out for investors (since the rating system should be more transparent)

Other than that, it’s difficult to predict that might happen. The new portfolio manager hasn’t been rolled out yet, so we don’t know exactly how much freedom we have when setting it up (such as, whether you can limit your investments by potential return or just by credit group or by whatever combination of other data.) Whichever way it goes, the new year will start with a lot of portfolio managers being created.

11 thoughts on “Impact of the new Bondora Rating system

  1. For those investors who have always invested automatically using portfolio managers very little will change. ——— What about the intress/profit which will decrease rapidly?

    1. I think that it was unreasonable to expect for the returns on Bondora to stay as high as they currently are. If the rates don’t drop then other sites will take away their business. As Bondora has entered the international market, that will impact their rates as well.
      Also, I really don’t think the returns/interest drop will be as big or as quick as some people fear.

    2. Haven’t yet done a thorough analyses of my portfolio based on the ratings, but seems that probably the rates in general will drop on a large portion of my portfolio, however, for the riskier parts they would increase and I will be able to better select whether I want more riskier or more lower risk loans in my portfolio.

      Whereas so far it has been rather random per individual loan, but has given me a kind of a mix of loans at certain risk levels when using certain investment criteria.

      Will have to look at the data in more depth, but at this point I’m not even entirely sure the return will drop that significantly initially. In the long-term, for sure though.

  2. Sure, returns will even out, but I’d look at lendingclub as an example here. They have similar credit scoring system which assesses risks and prices loans. They have also done for longer period and with a bigger team and there is still room to achieve above average returns with some analyses and strategic picking of investments.

    Reported returns that I’ve seen, differ from 6%-11%, which is almost a 2x difference. In Bondora’s case the segment of borrowers ranges even a lot more on the different risk levels, meaning there’ll likely be an opportunity for an even bigger difference in returns.

    However, that will probably require a lot more analyses and work than it used to so far to get an above average return.

    1. I’m looking forward to someone running serious analysis on the data set with the new credit ratings attached. (Tauri and I compared our portfolios briefly – while his has significantly higher ratings, I have significantly higher returns, so there is some value to the higher risk – higher return logic still even with the new ratings I think.)

      1. Too early to say anything about the risk-return part yet as the risk-based pricing hasn’t yet been implemented. Your portfolio has old prices.

        Another issue is that by investing into the A1000 or A700 etc groups, they still tended to have some sort of different proportions of different risk level loans. Whereas A1000 had less high risk loans than A700, the latter might have had a bigger proportion of higher risk loans. However, in that segment all loans were priced higher (even the low risk ones).

        In other words, you could just by good luck pick more less risky loans from a riskier group with the old system, but with risk pricing model (assuming the model is accurate) that can’t happen as much and results could be entirely different.

        Of course, in general there is always the rule that higher risk has higher rewards associated with it and at least in theory higher risk should be able to provide a higher return also with proper diversification. At least that’s what investors see even in LendingClub, so there is some merit to the thought.

        However, that still doesn’t mean that higher risk segments are for everyone.

  3. The only thing to do here is trust I guess. Searching for AA, A and B loans this morning gives 2 Finnish loan applications as a result. Both rated B. One of them is without any verification! How can unverified loan be B rated?

    1. Actually we don’t know what characteristics are important to use regarding Bondora Rating. My experience shows that there is rather little difference between verified and unverified loans in terms of verification method. So if verification method is insignificant variable / factor, then it is understandable why unverified loan got really high rating.

  4. I think there is one bigger problem than all of the above for investors such as myself. Let me explain: I don’t worry too much about data / return etc, I know I get a lot better rate than in a bank no matter what.
    Therefore I try to get as much money in the system as possible when i make a wire into bondora
    Since the new system went into place, I am on standby, almost no money is getting into loans and I have several thousands sitting in the account…
    I decided to check the “market” and well., before the change there were 4 pages of 100 loans a piece available..and now I have a single page and it seems like there are only 50 loans available…huge difference

  5. over 150 loans available today, i guess things are picking up.
    Interestingly, bondora rating of those 158 loans shows:
    4 in the “B” category and 5 in the “C” category
    everything else (95% of loans available) are D or less
    It seems new portfolio managers will be available next week based on Bondora’s latest email

Comments are closed.