Investor Column

Using the secondary market for above average returns on investment

wiseclerk

With the majority of my p2p lending investments I hold the loans I invest in to maturity. Observing the market over the years I have observed patterns on the secondary markets that can be used to actively trade loan parts with the hope to increase achieved yields and I sometimes tried these.

I never specialized in this, so I never used fully automated bots, but did in some cases use some automation (Selenium or third party browser plug-ins). This article is not meant to be a how-to guide giving concrete instructions, that readers could just follow, but rather a list of things to consider and look for, should investors want to start testing strategies on the secondary markets themselves.

This article does not name or link any specific platforms, as I believe the same patterns and chances could evolve on new p2p lending marketplaces, and can be used there, you just need to look for them. Nevertheless all of the examples are real examples observed at the past by me.

Required market environment

The marketplace needs to have a secondary market that allows investors to buy and sell loans parts at premiums and discounts. Ideally without charging any transaction fees, but some strategies can absorb moderate fees. Furthermore it is advantageous if specific individual loans can be bought or sold rather than just random from the market or the investor’s portfolio.

The effect of premiums on yield

One central piece to understanding why trading (or flipping) strategies can be highly attractive is the effect of even small premiums pocketed on the portfolio yield. Take an investor that invests into a 100 (whatever currency) loan part and sells that part for 100.40 after holding it for 5 days. That is only a 0.4% premium, but the annualized yield is 33.8%. There is a big IF to that annualized yield number – it is only meaningful if the investor can seamlessly reinvest all his money in similar trades without any interruptions. So cash drag (effects of not invested cash) are very important.

The psychology

One would think that investors would always use rationale when investing into loans. However, I feel that a significant number of investors show one of these behaviors and the question an investor with a trading strategy should pose himself, is whether he can use that to his own advantage:

  • Scarcity
    If there is little left / available of a loan, investors fear of missing out
  • Unusual collector’s passion
    P2P Lending investors are told everywhere that diversification is vital (I tell them too). But especially on platforms with few loans, it seems that some investors strive to have a COMPLETE collection of loans in their portfolio and in pursuing that aim pay higher than average prices to acquire loans they ‘miss’ in their portfolio
  • Herd behavior
    Happens for example when investor (over)react scared by a news bit on a loan. In p2p lending that mostly effects the selling side rather than buying
  • Investors overvalue the realization of small gains
    Many investors are very happy to sell at small premiums –it just gives them a sense a positive achievement, not realizing that the market conditions for this specific loan part would have allowed selling at a higher premium
  • Blended by high numbers
    High nominal interest rates, and high YTM (yield to maturity) figures displayed by the platform overly attract investors

Two main strategy approaches

Investors can
A) Invest on the primary market and then sell the loan later on the secondary market
or
B) Buy loan parts on the secondary market which they deem underpriced and then sell at a higher price. Be it buying at discount and selling at a lower discount, or already buying at premium and selling at an even higher premium.

While yields achievable in strategy B might be higher in percentage, this strategy is much harder to execute, as the competition will most likely use automated bots. Also the total market size for attractive loans will limit the scalability. I never tried strategy B on a larger scale myself. For example I was never found of buying already defaulted loans at a huge discount. Nevertheless I know of some investors that fare quite well buying defaulted loans and selling them at a lower discount, pocketing any payments and recovery that occur while they hold them as an additional bonus.

Following I will concentrate on strategy A) Invest on the primary market and then sell the loan later on the secondary market – as I have used this myself on several platforms over time.

One important point, is that market conditions change, usually good opportunities will stop working after a few months or weeks either because too many investors try to use them, or more general  the demand/supply ratio changes or the marketplace itself changes the rules how the market functions.

  1. First an investor will want to look how loan information is presented on the primary and secondary market. Especially what sorting and filtering mechanisms there are on the secondary market. It is highly desirable that the loans the investor wants to sell later, will be listed on top of the list of all loans on the secondary market with either the defaulted sorting, or with an obvious choice of filtering (e.g. sort by descending YTM)
  2. Understand the allocation mechanism on the primary market. How does the autoinvest feature work exactly? If there is no autoinvest, then are there chances to heighten the probability of investing in attractive new loans? Either through automation, or just because new loans are released at specific times?
  3. When is interest paid? Does it accrue for each of the day held, or does the investor holding the loan at the date of the interest payment gets full interest credited. This is important, because if in the example at the start of the article the investor not only makes a 0.40 capital gain but also collects interest for the 5 days he held the part, it will have a huge impact on yield
  4. Usually for this strategy longer duration loans are more attractive. This is simply because they will allow higher premiums without making the YTM value unattractive for the potential buyer
  5. Usually smaller loans are more attractive. As there will be less supply it will be more liquid on the secondary market and more sought (see collector’s passion). No rule without exception. On one platform the biggest loans were the most attractive to be invested in on the primary market for the trading strategy. Why? Because this platform used dutch autions to set the interest rate and the bigger the loan the less the interest rate would go down. And of course loans with higher interest rates could later be sold at higher premiums
  6. Usually the time span a trading investor wants to hold on to a loan part, will be as short as possible (days). However there might be patterns observed where it could be desirable to hold for longer time spans. For example if on a marketplace there are repeated alternations between lots of new loans and time without any new loan, it might make sense to hold the loan parts till there are no loans available on the primary market and only then offer them on the secondary market
  7. Strategies that allow to hold parts only at a time when the status of a loan cannot change can be attractive. For example there was a time when it was possible to invest into very high interest, extremely high risk loans on the primary market of a specific platform and sell them at a premium BEFORE the first loan payment was due. Most parts sold within days, and the ones that did not sell at a premium, could be sold at par shortly before the date of the first loan repayment (this strategy worked due to a combination of factors: sorting by YTM, blended by high numbers (see above), and platform design – instant selling of loans offered at PAR).

So why is all this possible. Mostly due to market inefficiencies and lack of transparency and experience. The marketplaces are young, selecting and evaluating loan parts on the secondary market is not an easy task. And on many marketplaces investor demand outstrips loan supply . Usually the yields achievable with trading strategies go down as marketplaces grow (but the volume that can be used in these strategies might grow over time).

The most prominent question is, if an investor can scale strategies he uses to a level that is worth the time invested and the inherent risks.

Feeding Creatures of Habit

wiseclerk

Many Fintech startups compete with banks and other incumbents by offering easy to use and attractive user interfaces. They appeal to users because they offer a modern packaging for processes and at the same time work hard to make these core processes they concentrated on more efficient  (unbundling).

User interfaces are an important aspect for p2p lending marketplaces too. While a very innovative user interface might have contributed in winning the investor, once he registered he not only wants an easy to use user interface he also likes constancy. Sounds paradox?

Fact is, that some p2p lending marketplaces are not that easy to use and offer complex functionality e.g. auctions, secondary markets with discounts and premiums. The investor spends considerable time to learn how to use the functions efficiently. Once he has mastered to efficiently use all the features and reports to achieve good results he will dislike any major changes the marketplace introduces since these force him to “relearn” his way around and render his previously acquired level of experience worthless.

I have experienced this several times on multiple marketplaces. But you don’t have to take my word for it, just look at a forum after a major redesign and you won’t have to search hard to find lots of investors venting their negative opinions rather strongly.

Now knowing that p2p investors are creatures of habit, what could a p2p lending marketplace do to ‘feed’ those? Freezing in standstill is not an option. Even the most conservative investor expects the marketplace to evolve and offer new features.

My suggestions are:

  • The platform should decide early on for a main navigation structure and stick with that. Changes and optimizations should subordinate to that structure and not change this main navigation
  • Development of new features should take wishes of investors into account (do surveys) but not be driven by them entirely
  • Test extensively before releasing. I am repeatedly surprised how many bugs there are in main features after a release (that is they show in main processes and not only in special constellations)
  • Measure, measure, measure. It is sensible to do A/B testing for all larger changes measuring the performance of previously defined KPIs (e.g. bounce rate). If the new version performs worse than the previous version the team should be brave enough to scrap the new developed version even if that means time and cost spent for developing it is lost without an output
  • If URLs change or cease to exist do an automatic redirect to the new URL
  • Even if the marketplace does not encourage the use of tools and automation, it should not ignore the fact that some investors will develop tools and workflows that helps them to speed up their monitoring and investment. The marketplace should consider how the changed impact and process might impact these. The very least that can be done, is to inform investors in advance of an upcoming major change.

Continue reading →

Which P2P Lending Marketplace Do You Recommend?

wiseclerk

I am often asked “Which p2p lending marketplace do you recommend?“. It is a natural question to ask for people that are familiar with the concept of p2p lending, but have not invested yet.

I feel hesitant to answer it with an outright recommendation for any one marketplace.

Sure I do have my preferred marketplace. Everybody has. But ask 10 different seasoned p2p lending investors and you might get at least 5 different answers. What is right for me, may not feel right for you. There can be no one size fits it all for p2p lending marketplaces. Interestingly as a sidenote investors seem to have less problems to agree why they dislike a platform – and they can also agree on ‘better’ platforms, you just don’t get consensus on the best platform.

What an investor prefers is influenced by his personality and past investment experiences. Investors differ in the expectation they tie to the investment, in risk appetite, in how they perceive and gauge risks. They may prefer a more actively managed investment or a passive investment style. Some enjoy auctions and elements that create competition for others factors like user interface might be a factor that lifts one platform over another.

That such a variety of different models has evolved and still prospers shows that they cater to an audience that is not homogeneous in their needs and wishes. One could argue that there is such a variety because it is a new field and everybody was just implementing ideas and experimenting and there were no role models, but that eventually the models will converge towards a best practise model. And I believe that is and will be happening, but only to a certain degree. Doing business over the internet allows marketplaces to deviate somewhat from the mass market and develop a style that fits a certain clientele easier than it would be for an offline financial offer because the economics of reaching out to and serving this clientele are different.

One entrepreneur recently told me ‘We are different, we just need 10% of the users to like us’ (sry if I rephrased that to much). My answer was ‘Just don’t be to different. Investors are conservative. Why scare 90% of your potential customers away’. I still believe in my answer, because I think it commercially makes sense. However it is minted by my past experience and my perception of the investor behaviour. So I actually want him to succeed in doing things VERY differently and making it as satisfying and enjoyable for those 10% he wants to be the perfect marketplace for.

What do I answer on the question?

At conferences or in other situations without much time, I usually suggest several marketplaces the investor might want to look into and point to my blog for more information.

If there is more time, I usually ask questions to try to find out what the person is looking for, what factors are important for him and what his past investment experience is. Then I tell which marketplaces do well on these factors and might in my opinion be a good match based on what I understood he is looking for. It still feels imperfect and uncomfortable for me sometimes. Maybe it is just a cultural thing, that most people are not comfortable in making recommendations how other people should invest money.

What would be the best answer?

I often think, the straightforward answer is ‘It depends‘. I have never given this answer. Even in situations when I am pressed for a very short answer.

Calculating Yield with XIRR

Guest

This is a guest post by German investor Martin R..

P2P Loan Yield

On most p2p platforms (all of mine except Ablrate and Estateguru) principal is paid back monthly during the loan term. The remaining principal decreases every month, the interests do so
accordingly. Inexperienced people are frequently confused by that – a loan over 100 EUR, a term of 5 years and an interest rate of 10% doesn’t yield a profit of 50 EUR, but roughly half of that.
When you think about it for a moment the reason is evident: On average, the capital was only lend for 2.5 years, a part of the debt was already paid back with the first instalment. In exchange, the instalment – as sum of interests and payback – stays the same for the whole running time – minor deviations can occur because of dues of the platform.

Which leads us to a good approximate formula: The obtained interest is about half as high as they would be for a fixed deposit with the same conditions. As already mentioned, the stated yield is still right, though. There are many websites to calculate instalments on the internet you can use to play that through.

Admittedly, such calculations made beforehand become useless if losses or early paybacks occur. And actually, they always occur. How is it possible to stay informed about the current yield in that case?

Mostly, the provider offers calculated ROI calues in the account overview. The shown figures are rarely particularly meaningful, though. Auxmoney for example displays values which
noticeably exceed the interest rate of the lent money – of course that is impossible. There are bookings being conducted wrongly and early paybacks are taken into account as earnings –
that has been happening systematically for years and was never addressed or fixed.

Two ways of calculating yield

In principal, you have to distinguish between already obtained yields ( this is the figure shown by most providers) and the total yield expected at the end of the running time.

The first figure is a good review of the past, but could only be realised if you sold all
your remaining loan parts for their remaining nominal value. Usually, no losses are being considered, not even the already failed repayments. This means the calculated yield is generally too optimistic.
A yield (XIRR, RTI) shown by Bondora or Omaraha of 25% or even more may not be technically wrong, but is not the whole truth either.

Of course, the expected total yield is currently not definite. After all, both future losses and payments due to defaults can significantly affect the yield, meaning the values can only be estimated.
Many refer to a worst-case-scenario when they fully depreciate all credits in defaults and depreciate 50% of all credits that are overdue. But not even that is the whole truth, because usually some of the loans that are current now will fail as well.

The XIRR-function

Thus, you won´t be able to avoid doing your own calculations. Admittedly, it is not possible to do those manually or with help from a calculator for a single loan part with irregular paybacks, let alone a large number of credits. Continue reading →

Crowdfunding – Dutch investors – where to go?!

Guest

This is a guest post by Dutch lawyer Coen Barneveld Binkhuysen (see full bio at the end of the article)

Crowdfunding is growing exponentially in the Netherlands. Although the Dutch market has not yet reached the astronomical levels of the United States and the United Kingdom, many people have heard about the phenomenon and are intrigued by this potential alternative investment opportunity. While the Dutch market speaks a lot about crowdfunding, it is less familiar with the term p2p-lending (it is commonly available though). As this article covers investments in loans, convertible subordinated loans and equity, I will use the general term crowdfunding instead of p2p-lending.

In the first 6 months of 2015, almost 50 million Euro was raised via crowdfunding, which is double the amount raised in 2014. There are over 80 crowdfunding platforms active in the Netherlands, which makes it difficult for potential investors to gain an overview of the viable available investment opportunities. This article provides a general overview of the most important platforms active in the Dutch market. Furthermore, I will discuss some relevant topics in relation to crowdfunding, such as: diversification options, costs, default risks, cash flow, types of investment and the added value of a properly managed crowdfunding platform.

Overview investment options

In general, crowdfunding platforms in the Netherlands offer the option to invest in loans, subordinated convertible loans and equity (besides donations and the purchase of products). Each of these different investment options has benefits and drawbacks in terms of cash flow, risk and the potential upside can vary significantly:

Loans provide a direct cash flow to the investor as loans are usually repaid in monthly instalments. Loans only have a limited potential upside, maximized at the offered interest rate. Due to the monthly repayments, the risk decreases every month. Most crowdfunding platforms determine the interest rate based on the envisaged risk. As far as I am aware, there are no platforms active in the Netherlands that provide the option to “bid” on loans in auctions.

Convertible subordinated loans (also called convertibles) are considered to entail more risk than normal loans as convertibles are subordinated to (normal) loans and other claims. Investors generally expect a higher return in exchange for a higher risk. Instead of offering a higher interest rate, companies issuing convertibles via crowdfunding offer the option to convert these loans into certificates of shares.[1] The option to convert may be restricted by certain conditions such as (i) a specific period in which conversion must take place and/or (ii) the condition that a sophisticated investor invests at least amount “X” during the term of the loan. For an investor it is important to identify any conversion conditions that may apply. If the loan is not converted into certificates of shares during its term, the investor will receive the principal plus interest payments at the end of the term of the loan. These investments might not be interesting for investors looking for a steady cash flow, but they can be interesting for those who want to have a shot at a serious return.

Equity is normally being offered in the form of certificates of shares (equal to the convertibles described above). Again, investing in equity does not create a steady cash flow for the investor. The terms and conditions related to the certificates of shares may (and normally will) restrict the option to sell them. Therefore, investors are expected to wait for the moment the entire company is being sold to an investor, which can take a long time. Investing in equity might only be interesting for investors looking for long-term investments. Then again, these investments do have the largest potential upside as the investor will profit from every increase in value once the company is being sold.

Balancing risks

Each investor takes, or at least should take, the risk of default into account, especially when investing in high-risk companies such as start-ups. Business cases of start-ups have not yet been properly tested and most do not, or hardly have, any financial buffers. Should the financed company go bankrupt, practice shows that only in rare cases (only part of) the loan can be recovered. Normally, preferred creditors such as banks and the tax authorities will receive the benefit of all assets left in the company and there is nothing left for others. Some platforms try to reduce the risk by requesting a personal guarantee of the entrepreneur, but this is of little use if the person does not have any assets.

The actual difference between investments in loans, convertibles and equity from a risk perspective is small. Investors having certificates of shares have a larger potential upside than the holders of loans. One could say that investors almost bear the same risk, but with different potential upsides. In my opinion the most important reasons to choose for normal loans are the fixed term and monthly repayments. If you are not in a hurry to make a profit and are going for the highest potential return, convertibles and equity might be a more interesting option.

Overview largest platforms in the Netherlands

After selecting the preferred investment instrument, it is important to select one or more of the available crowdfunding platforms. Without aiming to be complete, I list the largest and most active platforms active in the Netherlands below:

nl-geldvoorelkaarGeldvoorelkaar.nl is the national market leader and funded over 825 projects, with a total sum of over 66,000,000 Euro. The platform focusses on p2p-lending and only provides investors the opportunity to invest in loans. Interest rates range from 4% to 9% depending on the risk score determined by Geldvoorelkaar.nl. All loans are being repaid in monthly instalments as of the first month. By investing in projects via this platform, it is fairly easy to generate a decent cash flow. Up to now, 3.5% of my investments on the platform have defaulted. As the principal of one of the defaulted projects was almost fully paid back, my average ROI still accounts for about 6.5% per year. The other defaulted project was probably a case of bankruptcy fraud, which I expect to happen more often in the future. The platform opens several dozen new projects every week, which creates sufficient opportunities to diversify your portfolio and reinvest your money. An investor must pay a fee equal to 0.3% * loan duration (in years) * invested amount (which amount will be refunded if the project defaults).

nl-oneplanetcrowdOneplanetcrowd claims to be Europe’s leading sustainable crowdfunding platform. Since launching in 2012 it raised over € 6 million in funding for more than 100 projects. Oneplanetcrowd operates in Germany and the Netherlands and is planning to open in other European countries soon. It provides investors the option to invest in loans and convertibles (apart from donations and presale options) and offers some of the most interesting investment opportunities, such as Snappcar and Wakawaka Power. Various projects offer the opportunity to co-invest with sophisticated venture capital firms as these firms invest simultaneously with the crowdfunding campaign. In my opinion, this is a huge advantage for investors as VCs tend to do a thorough due diligence before choosing to invest. The platform only allows companies with a sustainable philosophy to start a campaign on the platform. Their goal is to provide high quality investments with a decent return to investors. Although this is a good niche market, the strategy makes diversification opportunities fairly difficult. Investors do not pay a fee on Oneplanetcrowd.

nl-other

KapitaalOpMaat and Collin Crowdfund are some of the main competitors of Geldvoorelkaar.nl as these platforms focus solely on loans with loan periods ranging from 6 up to 120 months and interest rates of 5.5% up to 9% depending on the calculated risk. Almost 6.5 million Euro and 13 million Euro have been funded via these platforms, respectively. Investors on KapitaalOpMaat pay a one-time transaction fee of 0.9% and a yearly fee of 0.85% on Collin Crowdfunding. Both platforms provide discounts to investors investing more than certain thresholds.

Bondora is a European platform offering the opportunity to invest in loans on a European level. Although this is by far the most sophisticated (international) platform available to Dutch investors, its presence is fairly unknown to most Dutch investors. Already more than 35 million Euro has been financed via Bondora. Investors are allowed to choose their own investments on the primary market, but most loans are filled in advance by a bot. Therefore, it will be necessary to invest automatically via the provided bot in order to obtain sufficient loans. This enables the investor to invest in literally thousands of loans differing in purpose, country and risk. All loans are repaid in monthly instalments on a virtual account. Bondora also offers the option to purchase/sell investments to other investors on its secondary market (with a premium/discount) against a fee of 1.5%. Investors do not pay any fees on the primary market. Although Bondora claims an average ROI of 18.75%, many investors complain about the large number of defaults. As the minimum investment is only 5 Euro, the threshold is low.

Symbid is one of the established Dutch crowdfunding platforms and focuses on equity (certificates of shares) and loans. Although Symbid seems to suggest that already more than 300 million Euro has been invested via their crowdfunding platform, the actual amount funded by the crowd is closer to 6 million Euro. One of the advantages of Symbid is that it offers the option to sell your equity to other investors on the platform. Continue reading →

Bondora Investments Using Decision Trees – Review of Progress – Part 6

Guest

This is part 5 of a series of guest posts by British Bondora p2p lending investor ‘ParisinGOC’. Please read part 1, part 2,  part 3 and part 4 and part 5 first.

Plan Your Change And Change Your Plan!

As stated in the previous article (see part 1-3) and revealed in the graphs of performance, I started using the Decision Trees in response to the rapid rise in defaults in my portfolio. Except for very small numbers of “opportunistic” purchases, I have maintained a strict discipline on purchase in order to ensure that my progress could be monitored and assessed. As my confidence has grown, I have modified this discipline to take advantage of the Bondora environment to achieve the demanding personal goals I had set myself when I first started. These included only purchasing Loan parts that should accrue 50% interest over the forecast life of the loan – i.e. should turn 5 Euro into 7.5 Euro over the original loan period.

Since early June, I have modified this discipline further and now purchase loans that, whilst still meeting my overarching rule of looking for 5% to 7% historical default levels, do not have a high enough interest rate to meet my earlier profitability goal. I intend to try and sell these loan parts on the Secondary Market with a short-term profit goal, after Purchase/Sale costs.

This further leg of my overall strategy is still in its infancy, but the results from my use of Decision Trees in my initial selection of Loan Applications suggest I am buying the best performing loans available. This means that should other investors not share this view, I will at least be left with Loan Parts that will perform well for me for the time I hold them.

Given the latest changes at Bondora mentioned earlier, if I can only acquire “good” (as defined by the Decision Tree analysis) from the Secondary Market, it may be that this buy-to-sell tactic may not be possible into the future.

Tree development

Tree Analysis

In the previous article (see part 1-3) on the construction of the Decision Trees, I explained how I had made adjustments to the overall analysis process to give more weight to factors such as “Total Income” in the actual Decision Tree analysis. I have kept the included data under constant Review and have added a few further fields to the analysis process, in particular the field showing the “Total Monthly Income/ New Repayment”. As stated in the first article, this needed to be modified from an infinitely variable value into 20 ranges, each of equal numbers of samples.

I mention this particular field as, since January 2015, it appears as an important feature in both the Estonia and Finland Trees and continues to appear more often in these Trees.

Volume and confidence

It is a fact that Estonia has been the largest market for Bondora from its days as Isepankur. In simple volume terms, the data I use (from 1/1/2013) shows that Estonia accounts for c.50% of the total loans, with Finland and Spain making up about 25% each. Slovakia is simply no longer mentioned in polite, Bondora society, so I will pretend it never happened!

Whilst it is true that Estonia has a lower historical default rate, in the dataset that I use, defaults do occur and are presently running at around 11.986% (1009 out of 8418), compared with exactly 18% (576 out of 3200) for Finland and 27.059% (1022 out of 3777) for Spain.

The above figures carry several implications as follows:

The Estonian Tree is fairly static with few changes at the highest levels. Estonian Loans within Bondora bring with them a richness in the data, by which I mean that the original Credit Scores are well represented across the Loan Applications compared to Finland and Spain, which are almost entirely populated with examples with a Credit Score of “1000”. What this means for Estonia is that the Decision Tree neatly shows that the Bondora Credit Score is relatively accurate, with higher numbers of defaults at lower Credit Scores. Thus it is that the historical record shows that Loan Applications with a Credit Score of “1000” (the highest and most sought after) make for good hunting when searching for segments having a default rate of less than 5%. Indeed, it is not uncommon for the Decision Trees to reveal segments of 50+ examples with NO defaults over the last 2.5 years.
Finland and Spain however, with very few historical Loan Applications with a Credit Score of anything other than “1000” combined with a default rate 50% and over 100% higher respectively than Estonia AND volumes less than half that of Estonia, provide pitifully few obvious segments with a sub-5% default rate AND sufficient numbers of examples to support anything like the confidence levels of Estonia.

I believe that the lack of richness in the Finnish and Spanish data is revealed in the overall structure of the different Trees.

Estonia

The top-most branch in the Estonian Tree is based upon the Employment Status of Estonian Applicants. This represents 5 different values: Full Employment (c.90%), Entrepeneur (c.4%), Self-Employed, Retired and, finally, Partially Employed (these last at c.2%).

The Credit Score generally appears at the 2nd, 3rd or 4th level below this and, as stated above, provides a firm “fault line” between >5% and <5% default rates in most of the segmentation below these levels.
As noted earlier, for those in Full Employment initially Income and latterly the ratio of cost to income (which I refer to subsequently as “Affordability”) is the next most significant differentiator followed by Credit Score with the paths exhibiting differing significant data elements somewhat below this level.

A strange (in my eyes) feature of what I call “Affordability” that appears in the Estonian Tree for those in Full Employment is an apparent truth that the more someone can afford to cover the cost of the loan, the less likely they actually do so and the more likely it is that default will occur! 17.333% (65 out of 375) of those in Full Employment who appear to be most able to afford their loans go on to default whereas only 6.54% (24 out of 367) of those in Full Employment showing the lowest affordability have defaulted. So it seems that, in Estonia, the higher the ability to pay, the less likely this is to occur!

Finland

The lack of richness in the Credit Scores provided by Bondora for Finnish (and Spanish) Loan Applicants is revealed, as the Credit Score is the primary determinant at the top level. This is, however an almost totally useless determinant as just over 98% (just under 98% for Spain) of all Finnish Loan Applications carry a Credit Score of “1000”. Below this level, Employment Status is the prime determinant, as in Estonia, but there any resemblance ends as lacking the Credit Score and with lower overall volumes and there is no common thread to the analysis.

Latterly the ratio of cost to income (what I have termed “Affordability”) has crept in at lower levels but there is no pattern to be discerned and the Tree has not settled down to any pattern at the lower levels with changes occurring at all iterations.

Such are the problems with low volumes and high default rates that I have changed the parameters for the Decision Trees for Finland and Spain to force the analysis to work with higher volumes in the nodes and leafs (end points) in an attempt to increase confidence levels. This has the unfortunate side effect of there being few leafs with a sub-5% default rate, the notable exception being a leaf of 23 examples with a 0% default rate.

Spain

As noted above, Spain shares with Finland the feature of Credit Score and Employment Status being the top 2 levels but for Spanish Loan Applicants in Full Employment, the number of Dependants appears to be the most important factor and has remained so for over 6 months of analysis. This data element does appear occasionally in both other trees, but only at much lower levels.

Other than this notable difference, the overriding feature of the Spanish Decision Tree is the lack of leafs showing a sub-5% default rate. Even where sub-5% default rates can be found, there are so few examples in the set with little in the way of trend or discernable pattern to support confidence at any instinctive level.

The best sub-5% default rate is a leaf of 21 examples, being 4.75%, for fully employed, divorced people with 1 dependant living in Pre-Furnished property! All other leafs with a sub-5% default rate are based on less than 10 examples. Many are only single examples.

A competent statistician (which I am not!) may be able to pry some hidden gems from this Tree, but I fear not.

Conclusion

The Decision Trees themselves, whilst changing over time, now appear to have settled down and changes that occur do so at finer levels of granularity with only occasional changes in the overall structure of any particular tree.

The numbers of samples (the complete Bondora dataset) entering the process have now reached the level where the Trees for Finland and Spain required modification of the actual Decision Tree analysis (known as an “ID3” tree) to increase the sample sizes at the lowest level. This has increased my confidence in the output even though the levels of default are so high that identifying sub-5% default levels leave me rejecting many more Loan Applications than I actually invest in.

My initial, restricted purchasing at the start of my new strategy has opened out over the course of period under review. After an initial period where my cash reserves grew to over 25% of my initial investment at Bondora, I am now confidently pursuing new avenues of activity with a view to maximising my returns within the opportunities suggested by the Decision Tree analysis.

This success in using manual selection of investment opportunities comes in the face of constant change at Bondora, change that is trying to move the investment process towards a passive, easy-to-use activity – an understandable business logic.

I take some comfort that my total efforts to date (which include aggressive management of non-performing loans) appear to be returning better than average results. In conclusion, I believe that my change from instinct- to numbers-lead investing has improved my portfolio performance when measured by this admittedly coarse scale of default level. Furthermore, this process has allowed me to start to take a wider view of the opportunities available on the Bondora platform and I hope to be steering my returns back to the levels that initially drew me to this platform.
In terms of the performance over the past 9 months, I experience severely reduced default levels going forward compared to those that triggered my realisation that a new investment strategy had to be formulated. I am now seeing levels similar to those last observed almost 2 years ago, on purchasing volumes approximately double those from that time. I will be the first to admit that the loans purchased over the last 9 months have yet to “mature” to the level of those from nearly 2 years ago, but I have a renewed confidence in the future performance of my portfolio at Bondora.

P2P-Banking.com thanks the author for sharing his experiences and strategy in detail.

Back in March an investor from Luxembourgh wrote an article sharing his experiences in applying machine learning to peer-to-peer lending at Bondora.