New portfolio manager Q&A with Pärtel Tomberg

The new portfolio manager for Bondora was released, so of course emotions as usual are running high. My first thought was – holy sh** – this can’t be how the new PM works – you just have to go through two confirmation screens and essentially pick one option out of three to start investing. So, after my initial reaction I had a chance to get some answers from the CEO Pärtel Tomberg about the reasons why the new system ended up as it did. (You can read their official pre-release post here.)

The answers by are by Pärtel (unedited) and my comments are in italics. (Before we start, I gotta say the answers I got from Pärtel were impressively detailed, and a much better show of communication then anything I’ve seen in the past couple of years. Some of the ideas I don’t agree with but he does show their mindset quite well.)

Screen Shot 2015-10-23 at 13.48.41

1. What was the design logic for having more simplicity? Don’t most investors want to more actively manage their investments? 

a.       Our main goal is to deliver net returns higher than available on the public stock markets regardless if you manage your portfolio actively or passively. We want to deliver the results regardless if you are investing 1,000 euros or 1,000,000 euros.

b.      Approximately 80% of the investments are coming from investors who only have used automated portfolios (passive investing) and 20% from investors who are actively managing their portfolios (loan picking, secondary markets, custom strategies on old Portfolio Manager etc.). We conducted extensive interviews and surveys in the beginning of the year and found out that the passive customers actually considered our product (Portfolio Manager) very complex and wanted a single control centre along with easy overview on top of getting access to Bondora’s returns. Active investors on the other hand wanted very granular reporting and loan picking options. It became evident that trying supporting both within Bondora website would be impossible as neither group’s needs would be satisfied.

(I have heard this complaint as well that the current PM as-is was difficult to use, and helped several people set their PM up.)

c.       Therefore we made a decision to build a super easy and light weight product for the retail investors and create an API to support the more active, trader types. We will now release an API that is open to third parties to build services on top of our platform (e.g.http://www.lendtower.com/, https://beeplus.me/). The second version of this API (version 1.0 is in our sandbox) will be coming early November and will include Primary Market buying, Secondary Market buying, Secondary Market selling and setup to support third party developers. Later this year we will also roll out full reporting package after we have updated and corrected the reporting package on the web.

(So far we haven’t gotten much info on the API and the things third parties are doing with it. I had two developers mark that at this point the API wasn’t as good as it could be, but hopefully Bondora is accepting recommendations from the people working on developing the interactions for API.

Also… corrected the reporting package on the web? I’ll believe it when I see it 😉 )

2. Why did you decide to change the way account value was shown ? (meaning the 60+ loans being counted only as the unpaid-by-this-moment as opposed to the whole value of 60+ with maybe a combination of expected % of recovery?)

a.       The objective of the dashboard is to give investors an immediate view of the profit of their portfolio. This profit is calculated using the net investments made by the investor to date (deposits less withdrawals) and the value of their portfolio.

b.      Investments into loan portfolios are fixed income investments whereby an investor exchanges a certain amount of capital for a steady stream of monthly payments that are higher than the initial investment. The key here is the monthly payments. Part of this return comes from interest and part of this comes from reinvestments interest that in turn generates both new principal payments and interest – e.g. cumulative interest.

c.       All loans are priced on the assumption that a certain proportion of loans are not repaid. The default and corresponding recovery is factored into the interest rate so that interest payments on performing loans are high enough so the total portfolio delivers the expected return.

d.      As interest and reinvestments of repaid principal are made monthly then all provisions are matched to the same period. Otherwise we would compare lifetime losses of a portfolio with the interest income of a a couple of months. This logic would potentially work only with a very short investment focus but as investing on Bondora is a long-term investment then it does not make sense.

e.       There are two financially sound alternative views to calculating the portfolio value the way we do in the new dashboard. First is based on IFRS rules for banks and the second based on repricing of all loans in the portfolio according to a similar logic we use for pricing new loans. We plan to roll out both methodologies by the end of the year so that in the future investors can choose which context fits their requirements.

f.       In case you disagree with our logic and would want a simple alternative to check the net annualized return than simply compare the interest income of your portfolio with the amount of capital you have invested on Bondora (deposits – withdrawals) for a certain period. This would be ROCE logic (http://www.investopedia.com/terms/r/roce.asp).

g.      PS! Discounting the balance of overdue loans and matching this number with income makes sense for mortgage and super-prime loans (e.g. ZOPA) where the income of the performing portfolio is negligible so a loss can effectively never be priced/repaid by income on the portfolio.

(I think this is a tricky issue – overall for new investors seeing a more balanced view, but long term this becomes a bit problematic because the number of what’s shown on the dashboard vs what’s shown in the pie chart will differ by a magnitude of 10x.)

3. How should the passive investor decide between the three risk classes? (What should be the ‘decision tree’ that leads them to a specific type of investment?)

a.       First of all an investor should decide if and how much of their portfolio they would like to invest on Bondora. Passive investors typically invest between 5%-10% of their portfolio on Bondora and are not that much concerned of additional diversification (as they have already diversified by only allocating 5-10% to Bondora).

b.      The three risk-return options allow you to decide if you want to build a portfolio that is on average less profitable (and risky), more profitable or at average profitability.  

c.       Each of these options comes with different expected returns however the actual risk of the portfolio (losing money on a portfolio vs. the gains) is always determined by the diversification level. Our statistics has shown that after 200 investments the volatility of returns stabilizes based on if you took the less or more risky route. However in most cases investors earn considerably more than anywhere else on the market and almost never lose money (enclosed).

d.      In summary, investors should first decide if they want to by the types of unsecured personal loans that Bondora can deliver and how much they want to invest. If this amount is enough to diversify (we think that you should consider building up a total portfolio of at least 1,000 euro to meet this condition) then thereafter the actual risk of your portfolio is not influenced anymore by the types of loans you pick.

(Once again… the decision making is slightly problematic. When I try to think about which of the three options I should choose, I would probably struggle. Long term valuations of high risk – high reward can be hard to project the value into the present.)

4. Do you feel that the current one-window check screen for the investor to verify their understanding of risk is sufficient to guarantee that investors are adequately analysing the risks associated with investing? (How are less knowledgeable investors protected from making bad investment decisions?)

a.      In order to invest on Bondora the investor needs to review the information on our landing pages that based on FCA guidelines include full information about risks and processes. You should be confident that you want to invest through Bondora when you sign up and you should only invest a certain share of your portfolio. Typically our customers invest 5-10% of their capital with Bondora and they consider this part of their alternative investments portfolio (potentially higher return but less liquid/higher risk).

b.      The Portfolio Manager page includes information on the expected returns (that are net of losses), an overview of how the portfolio is expected to grow, the expected composition of your portfolio (by risk level and country), expected numbers based on levels of diversification as well as detailed explanations of how different numbers are calculated.

c.       We believe all these steps and detailed information is sufficient to make an investment decision on a 5-10% proportion of an investors portfolio. This level of detail is considerably more extensive than you would find with the people selling pension funds (in supermarkets), online FX sites or online asset managers. We also think our information is considerably more detailed than any other marketplace lender.

(The last in particular is an interesting statement, that your P2P portfolio should be 5-10%. While most bloggers try to direct most investors into this direction, then not many starter investors follow this – for many the P2P section of their portfolio is anything from 50-100%.)

5.  How does the risk balancing process work? How far “out of balance” will the portfolio tip before you stop investing into a group to wait for others to rebalance?

a.       The system calculates your portfolios risk level after each batch of investment and only buys batches of loans with a risk level lower or higher than your existing portfolio depending on your strategy. In general customers who have signed up for the new portfolio manager are looking to decrease the risk of their portfolio.

(This would definitely be interesting to see. If this algorithm works well, then whoever designed it definitely deserves a raise. Also, when the previous PM was launched in January I also set it to be more conservative than my existing portfolio, so I guess I fall in to that ‘general’ subset of investors.)

6. The sizes of the loan pieces – take the sample of my portfolio, what would the investment per loan be (and why isn’t this visualised anywhere?)

a.       The size of each investment is determined by the number of borrowers in your portfolio. The investment size is doubled after each 200 investments (the file enclosed earlier explains the logic). Your (as in me, Kristi’s) portfolio has 680 borrowers which means that your bid size will be 40 euro (or 0.72% of your outstanding portfolio). This level delivers an optimal investment time as well as risk diversification.

b.      This number is not visualized as the people we interviewed when developing this product never raised the topic. We are happy to add this in case our customers request it. Most people never think of how much they would invest into granular assets within the category they are investing in. For example if you choose a pension fund you never think or decide on how many euro per month is invested for each stock or bond in the portfolio.

(Yes, please visualise this :) )

c.       If everyone would sign up to the new product the roughly 83% of the investors would be at 5 euro investment size however this number is more equally spread when we look at the amount invested.

Screen Shot 2015-10-23 at 23.23.50

(Alright, so now this is probably the most controversial of all the issues. My current portfolio value as you see above is about 6,5K. This means that at the rate I invest, which is about 150 euros per month, with returned money reinvested I will be investing into a total of…. 6-8 loans per month. I’m not entirely sure how I feel about that, especially since I’m on track to get to the 80 euro pieces quite soon. On the other hand, I was super surprised at the data that so many of the investors are at such small portfolio sizes that they would remain at the 5 euro piece – this means that they have <200 loan pieces, meaning a portfolio of <1000 euros. The data shows an intriguing look into the investor base of Bondora – a lot of retail investors who invest small sums, and a small % of investors with significant portfolios that make up almost a third of the money invested. This is the part where you have to keep in mind though – the people complaining on the forums are those who are a very vocal minority – the majority just sits still and lets their money get invested without causing a hassle on the forums.)

7.  Will the 80 euros per limit also apply when investing via API? Why was the decision made to lower the investment amount per loan? How many portfolios are likely to be investing at the max level of 80 euros per loan? (Also, if your account has less money than the assigned ‘loan piece size’ then will it still invest or wait for the money to stack up aka the problem of ‘cash drag’?)

a.       The 80 euro per loan limit was set to ensure that the proportion of each new loan in the portfolio is not higher than 1% of the deployed balance in most cases for customers converting from old products to the new. Such restrictions do not apply when investing over the API – you may fund the entire loan if you want through that channel. As the data showed earlier – roughly 2.6% of investors would be investing at that level in case everyone would convert to the new product.

b.      The new Portfolio Manager never invests fractional amounts to ensure the diversification rules are properly applied. This is already the case for most investors as the minimum investment amount is set already to 5 euro.

(I essentially read this as – the people who would use the API or those who would be investing the max 80 euros per loan piece are likely to overlap anyways so this is somewhat of a non issue.)

8. Does account value function imply that at one point selling whole portfolios will be possible? (Since 60+ loans are now sellable on the secondary market already.)

a.      The new Portfolio Manager will in the future include the option to sell part of or entire portfolio. Other investors using the Portfolio Manager will then pick these up in case there are not enough loans on the primary market.

b.      We are currently working on building the models to price outstanding loans so that loans on the secondary market would be priced according to the same logic as on the primary market. In essence this would mean that certain loans would be sold at balance value, some above and some below.

9.      How many investors are you predicting will be using the passive web interface vs how many will be holing out to use the API?

a.      We expect that roughly 80% of investors will use the passive web interface and rest will come through the API. Already after two days the new Portfolio Manager accounts for 45% of the new investments without any marketing.

(I think this % is very high – it’s a bit scary to seen that many people be so passive with their investments, but overall, Bondora is just about as passive for them as an average investment fund now, so maybe they shouldn’t be judged that strongly based on that. I’m tempted to try the new PM myself as well, just until the API is live since it will take time to test most third party apps for that as well.)

10.  Many of these changes are likely a precursor to more institutional investors coming on board – are you running any scenarios to assess the likely future balance between retail & institutional investors?

a.       We are building a structure where institutional and retail capital is collectively combined to allow consumers move away from banks. It is certain that institutional capital is going to be higher than ’self-directed’ retail capital as the markets are at very different sizes. However you should also understand that raising money from banks (who hold your deposits), pension funds (who manager your retirement plan) or insurance companies (who manage the funds to keep your home safe) are all retail investors simply packaged together.

(While this is already long then some final comments –  1. I do think that most people underestimate just how passive investors are, and a lot of our first impressions are strongly coloured by bad decisions made in the past by Bondora. While talking to Pärtel then he also mentioned that some reorganising is happening to manage investor relationships better. Though they have done this before, then I hope they do it better this time to generate answers for the more complex questions that investors are likely to ask. 2. If you feel like the new PM isn’t for you, then I recommend against activating it – but as you see from the data above most people are likely to activate and just be silent about it. 3. I reserve the right to guess more about these things once they’e been live for long and we can see a bigger impact of this on the way the market works.)

To finish up, a XIRR chart, kindly shared by Pärtel, to demonstrate the need for diversification:

Screen Shot 2015-10-23 at 23.59.48

23 thoughts on “New portfolio manager Q&A with Pärtel Tomberg

  1. Great interview. Thanks.

    I would like to add a comment to the obviously silent majority who appear to enjoy the fuits of Bondora. MoM loan growth (excluding all cancelled loans) of Bondora is flat for over a year now and in September 15 it went down a lot. I believe that this is a very clear sign of not having the convidence of the existing imvestor base or not attracting enough borrowers. I am not sure yet how this month will look like. But not growing anymore is a worrying sign for a start-up.

    Bondora now faces a lot of competition from other plarforms.

    1. Yes, and looking at where the growth should come from – those are that 5% of investors who probably want more info & great tools. There is only so many loans that the people who invest 5 euros per loan can finance.

      1. Do you really think growth is limited by available capital? I really think growth is limited by finding viable borrowers and lacking further expansion into other markets and seeing increased competition I think ironically the reduced volume is a good sign because it means that Bondora doesn’t just push up the volume by accepting lower quality loans like so many others do.
        Whether it’s sustainable is of course a different question.

        1. I think growth is limited by people’s willingness to invest the mentioned capital. We know that people have the money – but if they do not trust the investment then they will not invest it. People don’t seem to have enough faith to allow for growth atm.

          1. Really, that would be a quite unique issue.
            Do you have data? What is the percentage of loan applications that doesn’t get funded on Bondora and has it increased recently?

            1. I mean, they recently changed the way loan applications get filled, since loan applications *were* going unfilled. I assume there is data about this somewhere, but it’s probably one half the problem, lacking lenders is probably an issue as well.

              1. Well, that’s what I meant with “unique”. If this is the case they are probably the only ones with that problem. All other platforms currently get flooded with private and institutional investments, with commitments sometimes one or two magnitudes higher than actual closed loans.
                When this whole P2P thing started 9 years ago or so we all thought that the biggest problem was getting people to invest their money but this hasn’t happened. At least since the financial crisis investors were very quick to jump on the train (partly due to the lack of alternatives). But the credit market is highly competitive, most markets have established banking infrastructure that can compete and fight with low prices and also in most markets outside of the UK the lack of rating experience has eaten the cost advantages P2P loans have over traditional credit so it’s a tough fight for the borrowers.
                Investors, nah, plenty a dozen.

  2. Great article Kristi, I was waiting for you to post something on the new Portfolio Manager :)

    I like a more aggressive lending strategy, though still diversified. I’m not expecting to use any of the money for 20 years +. I had the old PM setup to:

    20% HR
    20% F
    20% E
    20% D
    12 % C
    6% B
    2% A

    What I don’t like about the new PM is, that now I’m forced to use Progressive, as the most risky option. And the Progressive is 57,39% E and 42,61 F loans. Why this setup of only 2 risk categories?

    Unfortunately I only have ~1000 EUR left in the old PM, and after that I will be forced to do 100% manual picking or start using the new one. As I transfer at least 1000 EUR / month, that reality will hit me within 2-3 weeks.

    I love the setup of the new PM, that it auto-adjusts and it automatically chooses the amount per loan. But if I in theory wanted 100% HR loans, why shouldn’t I be able to do that in the web interface?

    I’m really looking forward to seeing the API options that will be available to us later on. I’m not a programmer so I’ll have to buy/use whatever option that becomes available from developers.

    1. Well, so far no one has shared any tools they’ve created for the API, so I’m also waiting. I mean, there is a certain level of trust though that you should have for whoever develops something for you to use on your investments.

      But yes, all of the alternative setups like 100%HR would mean using the API. Many investors will have to either wait a while for that to go live though or to use the other PM for a while.

      1. The new PM is now live but their API is far from being completed yet… So don’t hold your breath.

        I plan to write my own software to use the API. Maybe in a form of webpage that is usable to everybody… If you are familiar with Excel and for example would like to create kilometers long formulas to eventually find out the decision whether you would like to invest into that particular loan, then you probably want to use something i’m going to make. :)

        Thankfully i have made my PM with portfolio size 100 million. I just made AA-C, each with 25% (but it doesn’t matter, because the portfolio size is set to “infinite”) and so whenever appropriate loan comes in and i have the money, my PM will invest. We’ll see, how long they keep old PMs working. 😀

  3. OK, I’ve already posted this on facebook but maybe the blog is the better place for the comment plus I have two additional issues with Pärtel’s comments and a question…. In number 5 about taxes, I think if you have a chance to ask Pärtel that, too, it would be quite an important aspect to even make this whole scheme work at all for a lot of investors.

    First of all, thanks, Kristi, that’s a really interesting article.

    Actually I can quite well understand that many people want to invest passively and if it works – and with “it works” I mean it creates the return predicted by Bondora – I see no reason why one would not want to use it.
    But I’m having a few issues with how Bondora handles this and with Pärtel’s answers:

    1. His proposition to always have better returns than the stock market. In an efficient market – and competition in the P2P sector is picking up quickly – that can’t work so if you keep the goal it drives you towards too much risk

    2. Passive investing means trust and trust means transparency. Traditionally Bondora has been pretty good at this but this “we plan to put a value on overdues” thing has a very high risk for transparency and that’s due to the delay you get for realization of recoveries. A recovery on average will take more than a year so whatever you see in the statistics today will be old and if things start to go south your PM’s algorithm would only START to react a year later. That’s OK for a long term stable business but for something as quickly developing as P2P credit markets – puh.

    3. Spanish loans. And this is the big issue for me: they use this to sell credits that they KNOW are not profitable. By their own statistics Spanish loans don’t create value for the investor, on average the investors lose money. They create value for Bondora through fees. So they expect me to invest through an automated vehicle that knowingly invests part of my investment into a lossy asset category just to create fees for Bondora. I will not accept that and that’s why this PM is not for me.

    4. Pärtel’s comment on the overdue vs. value thing in 2d. I mean…. His own argument is that you should compare Apples to Apples but in fact they are comparing an accumulated value (outstanding principal) with a monthly one (overdue principal) and that simply doesn’t make any sense. It would make sense to compare overdue principal with repaid principal (in the same month) or even with cash flow. But if they compare to portfolio value or accumulated, outstanding principal they have to compare to the value of the overdue principal everything else is just misleading investors. I really get a bad feeling whenever they do one of these changes that look like they try to hide some troth and unfortunately they are doing a lot of them recently (reschedules, changed statistics…)

    5. Bondora is going for international investors but this whole procedure completely ignores tax regimes. Investors for Germany and Austria for example can’t deduct losses so the whole argument falls apart. This is especially a problem for valuation of overdue loans. I would get a lot of loans that would produce huge losses for me but I can’t deduct those so I can’t just count them agains interest to get a real return figure, I would have to deduct taxes on _all_ interest earned and _then_ deduct losses.
    Now, while this is a German/Austrian speciality other countries as well might require loss certificates for the investor to be able to make a claim against the income, this is for example true for the US, I believe, where Bondora seems to be wanting to expand to. Since I don’t see Bondora officially being able to certify anything the tax problem applies to these countries as well. What will Bondora do about this?

    Thanks again

    Joerg

  4. “On the other hand, I was super surprised at the data that so many of the investors are at such small portfolio sizes that they would remain at the 5 euro piece – this means that they have <200 loan pieces, meaning a portfolio of <1000 euros."

    or the fact that 83% of investors would be in the 5 euro section could mean that most investors are badly diversivied. because someone who invested € 10.000 in 100 borrowers would still be in the € 5 section, wouldnt s/he?

    1. True, that is a good point. I would like to hope though, that most people’s portfolios are just small enough, not just terribly diversified. (But then again, I’m sure that there are people who have invested 10K into 100 loans…)

        1. In addition to lack of diversification it could also be an issue of strategy (for example the first set of Finnish loans were terrible).

      1. You’d be amazed how many people have e-mailed me about what’s wrong with their portfolio and then it turns out they have the typical €5, €5, €5, €500, €5, €5… type of diversification in their portfolio.

        There’s a reason why I’ve highlighted this several times in the list of typical mistakes P2P investors make 😉

        Of course, I doubt it’s the biggest portion in that segment there, but it is a problem anyway. For some reason or another there are still people who find it a good idea to search for hours for a single loan to put all their €100 “portfolio” into.

  5. Totally agree with other investors, this new PM is made to complete all Spanish loans as well as HR and F categories loan which are not enough invested in. It is a way to force investors to invest in loans they don’t want to. It is terrible.

    Otherwise, all my loans are 5E, even if my portoflio is 10kE.too badly diversified ? Not sure. so many overdue loans and recovery process is infinite, better to wait 1 week for more laons than 2 years recovery process.

    That is just my opinion.

    Thanks for your interview and your blog

    Arnaud

  6. Hallo Kristi,

    thank you for the interview.

    Please, write an article on API for Bondora.

    I am one of investors on Bondora who invest manualy and I can not accept PM in actual form. I can not turn PM on.

    It is in Bondora interest to sell as much loans as possible. It does not matter what quality of loan it is till there is enough people with PM turned on… .
    I accept that, but I simply prevent to be the one who they sell bad loans to. And so will 10 % of investors controling 57 % of capital investing on Bondora.

    I believe Bondora will soon offer solution for theese investors so Bondora will not lose oportunity to get their capital invested on Bondora.

    We will see how this will evolve.

    Stano

Comments are closed.