Getting started - or not.

Not quite out yet but exiting as fast as I can

%age is probably the wrong solution (haven't written the algorithm yet, that's on next weeks to do list) so more likely we'll give a small boost to people with, say, less than £100 invested to help new lenders get funds deployed but otherwise it'll be equal opportunity on that factor.  Length of time waiting is also liable to be riggable or to cause issue when amounts are changed so is likely to be a tiny factor.

It is very important for us to balance the needs of smaller investors alongside those with deeper pockets and more to invest.  Both need to get their capital away and I'm determined to build a system that isn't riggable by setting large targets or using weight of capital to get preferential treatment.  The problem we face is that with limited supply it will *always* be impossible to come up with a fair system that works for all.  Hence this being part of a wider spectrum of changes to our offering intended to rebalance supply and demand, amongst many other goals.

Not everyone is going to like what we do, simple fact of life especially if there's a group who feel they currently have an unfair advantage and we take that away, but I'm personally determined not create an impenetrable black box that no one understands and that fails to deploy cash for anyone.  I'll always be happy to explain our algorithms and procedures on this board and will always welcome feedback if it helps us achieve fairness.  But to a degree whatever solution I come up with is a slave to the business originating enough loans to keep lenders happy.

A few points from my experience to date:

1. I started 6 weeks ago and through bids and AI have deployed my capital over 15 loans giving me decent diversification at this stage. AI works well in my view.
2. As to how AI should allocate available loans across the lending community keeping it simple comes top of my list. A simple queuing system and available funds is enough.
3. There is always a danger that in trying to solve a current perceived problem and to keep everyone happy that the output is a complex hard to understand solution which has 
    the opposite effect.
4. If demand outstrips supply then some lenders will always be left short. Focus on fixing that balance for the best results.

All opinions welcome.

Mike

Surely the fairest way is to have a rotating queue. When loan parts are put up for sale, the lender at the front of the queue buys one part and then goes to the back (or falls out of the queue if they have satisfied their mandate). New mandates are also added to the back. in low supply situations, the available parts are divided equally among the people who've waited longest, in high supply situations everyone gets what they wanted. IMHO, any attempt to allow anyone to jump the queue in order to "make it fairer", actually does the exact opposite.

How do you determine order when mandates are placed in bulk via an API or other automated system, what determines their start order?  How does this work with loan units being broken up into smaller parts e.g. £100 loan unit is broken into a £20 chunk and an £80 chunk as the first person in the queue wanted £20.  Does the next person in the queue get the £80 chunk and get shuffled to the back of the queue?  Or does the £80 chunk get shuffled to the back of the list of loan units awaiting sale so that a £100 chunk can be brought forward for the next person?  If that system is used then it works fine when large denominations of loan units are sold in bulk, but what if we're dealing with small ad hoc sales?  In a heavily over-subscribed environment would a new lender to the platform simply go to the back of the queue on all loans and take forever before they start being able to deploy any cash?  Should someone who just wants blind diversification be able to deploy their cash more rapidly than someone who wants to hand pick a few loans in which to lend, as they would with your solution.  And so on.

I'm all in favour of simple, but we don't have a simple system with one simple standard set of behaviours.  The simple answer is to increase supply, and that is certainly plan a, but the system needs to be coded to cope in any scenario.

My intention is that if you want to invest £100 you have to have £100 in your account, no more no less, and that it should deploy as quickly as possible.  My challenge is to make that the case whether you want to invest £10 or £10m, but that is what I'm striving for whilst trying to find a way in which the system cannot be gamed.

Contrary to other platforms I won't pretend that our solution is perfect and I'll be on this forum to justify choices, explain how it works, and listen to feedback on how to improve it for all.  I won't be able to please everyone, you never can, but hopefully I can find a way to balance the needs of the many.

It's part of a wider vision for the site, one that was always coming with AI being a small piece of that puzzle.