Loanbase, one of the peer-to-peer (p2p) bitcoin lending platforms announced today that they will soon unveil a new auto-invest feature. This comes as good news to the bitcoin p2p community, but the change could have some drawbacks for certain users. One of the differences between some of the main p2p lending platforms is in the way they price their loans. For the leading platform BTCJam, borrowers are assigned an interest rate for their prospective loans based on how the borrower rates according to BTCJam’s credit rating system. The benefit of this loan pricing format is it takes most of the onus off the lender. The lender doesn’t have to do their own diligence, they can just rely on BTCJam to price the loans accurately. It therefore makes it easier for novice “investors” to participate in the BTCJam p2p community without having to build much of a investment policy. For example, if a BTCJam investor knows that the system has assigned a loan a specific interest rate based on the risk criteria as set out by the BTCJam credit rating system, the investor no longer has to bother researching individual credits. Under the BTCJam system of risk pricing, the investor can indiscriminately invest in a large number of loans, and so long as those loans are randomly chosen, they will provide the required diversification to earn a profit. BTCJam is becoming the rater of individual credits. This system makes it possible to offer an auto invest feature because credit rating doesn’t matter as much.
The problem with this “risk based pricing” model is that since almost all credits on the bitcoin p2p lending platforms are high risk and low quality, the prices charged must always be high. The system does not allow a good credit to post their own price. The system of risk based pricing as implemented by BTCJam excludes good credits and makes the platform one for bad credits only. For investors, returns on diversified portfolios will look like the returns on a payday lending portfolio.
The difference of Loanbase to BTCJam and Bitbond was that borrowers on Loanbase could negotiate their own rate. It was up to lenders to decide whether this rate reflected the estimated risk of the loan itself. My suspicion is that this system was too sophisticated for the bitcoin p2p lending community at this stage of development. The platforms can probably drive more liquidity by using the autoinvest feature. Casual lenders on Loanbase probably found that their loans were having a default rate greater than what they were earning in interest and so casual lenders were loosing money. For sophisticated lenders, the Loanbase system was better because sophisticated lenders know how to calculate an APR and only invest in loans that had APRs that considered the likely default risk. Since Loanbase allowed lenders to negotiate a rate, I would often change the rate to meet my lending criteria and most of the time the borrower would accept my negotiated price. The data is available from BTCJam and Loanbase about default rates, lenders need to charge an APR around 50% or more because more than 40% of loans (of all ratings) will default.
My own Loanbase portfolio is outperforming my friend’s BTCJam auto-invest portfolios because I’m taking advantage of loans where the borrow is mis-pricing their own loans, since my knowledge of Loanbase credit risk is better than most borrower’s. Specifically with short term loans, sometimes the fixed rate would be 8% (which doesn’t sound so bad) but since it was a loan of 90 or less, the APR was in the many hundreds of percent.
Now that Loanbase will move to a risk base pricing model, the advantage of the sophisticated user will largely disappear since loan pricing will no longer be negotiated. This is too bad in my opinion because it will limit the ability of good credits to raise funds. Along this vain, I have been finding the lowest rated credits on Loanbase have similar rates of return to A and B credits since the lowest rating (E) included all new borrowers and borrowers who were otherwise less risky, but that included borrowers who were unwilling to disclose additional personal information (which kept them in the lowest rating category).
Ponzi schemes are also a key risk on these anonymous p2p platforms. Consider the borrower who plans to take out a loan on a p2p site and then abscond. At the lowest rating, the site limits the size that they can borrow, but as their rating increases with more personal disclosure and the repayment of past loans, they can borrow larger amounts. The ponzi operator will use the lowest rating for their first few loans, and make sure they pay those loans back. As successful repayment enables them to increase the amount they can borrow with a higher rating, they will default at a credit rating higher than the lowest. I have found that the lowest credits may have a better repayment history because those borrowers are trying to increase their rating (whether because they are honest or because they are planning a ponzi is irrelevant to the lender).