Who Pays What?

[PatrickHarnett]

OP updated.  Some additions, some shuffling.  Newly published credit rating metrics for Ineedausername to go with his insured deposits and uninsured PPT.

[1713558711]

1713558711

[1713558711]

1713558711

[1713558711]

1713558711

[1713558711]

1713558711

[1713558711]

1713558711

[payb.tc]

^^ patrick, did you update it from an old copy? i'm fairly sure you had already adjusted my minimum to 5 BTC but now it shows as being back to 10 again.

[PatrickHarnett]

^^ patrick, did you update it from an old copy? i'm fairly sure you had already adjusted my minimum to 5 BTC but now it shows as being back to 10 again.

Sorry, my mistake - re-fixed.  I have an off-line record that I use to generate the tables (otherwise they would look much worse) and didn't change it when I did the quick change last time.

[PatrickHarnett]

Kluge credit rating added.  AA

Despite the (insert bad word) things that occurred today, BDK and Kluge still obtains a very respectable rating.

[ripper234]

(insert bad word) things that occurred today

?

[PatrickHarnett]

(insert bad word) things that occurred today

?

Check the BDK threads.  Kluge had his GLBSE account plundered/screwed despite changing a password - an open session from a hack resulted in some fraudulent trades.

[Willowbitcoin]

Hi Patrick,

Are these lenders supplying you with identification (drivers license? etc ) to obtain these high credit ratings.

I understand the reluctance of wanting to give a "stranger" on the net a scan of my drivers license but on the flip side there must be some exchange of details when we are dealing with tens of thousands of dollars.

I feel identification should be a highly weighted metric contributing to their final credit rating.

Without the supply of this identification they are just facelesss names on the net and i would suggest that people are capped at a A rating or lower without supplying this identification.

I imagine you personally wont have a problem with it as your easily identifiable even to the point of your place of work.
Can we supply these same conditions to other lenders ?

I think it would go along way to increasing the profitability of their business to the level you are currently enjoying as depositors would feel allot more comfortable investing their dollars.


just my 0.2 cents

[Willowbitcoin]

Also... fwiw, you should add kludge to deposit takers as he also takes deposits.

[PatrickHarnett]

Also... fwiw, you should add kludge to deposit takers as he also takes deposits.

Yes, I will update.  I see his revamped (rainbow) OP has his CDs listed again.

[Willowbitcoin]

any comment on identification requirements patrick ?

Seeing as you are championing this thread, you would need to lead the call if you felt it would help secure lenders and borrowers.

[PatrickHarnett]

any comment on identification requirements patrick ?

Seeing as you are championing this thread, you would need to lead the call if you felt it would help secure lenders and borrowers.

I agree that ID requirements would be good, as would people providing additional information about their business models and supporting assets (credit ratings).  However, this is voluntary and I can not force anyone to do that.  However, if people make it clear to different deposit takers the level of disclosure they need, then that will improve this section.

In a similar vein, I am often asked for recommendations, but that is a bit more complicated.  Not sure what the general consensus is.

[PatrickHarnett]

While working with RustyRyan today it became obvious that start-ups present an interesting issue: with no deposits or liabilities, they tend not to produce meaningful metrics.  Therefore, having a minimum size makes sense. 

This ties into some of the other size dimensions that are relevant here, and that is the size of the deposit-taking operation.  For example, you might be looking at a deposit, and someone with 100 coins from one person might have a different profile to someone with 100 deposits for the same value.  Similarly, someone with 5000 coins under management might be seen as having a more stable profile, maybe not. 

In any event, collecting data on the size of business (total and # of customers) is relevant for some people, just the same way bad-debts are not relevant for those that do not provide loans (and might artificially increase a rating). 

• Possible size breaks would be 100/500/1000/2500/5000/10000.
• Possible customer number levels could at 10 and double 10/20/40/80/160

This might change the ratings slightly, but the idea is posted for feed back from whichever side you're on.

[PatrickHarnett]

And, to herald the other change - when this kicked off it was based around a deposit taking/lending coin model.  Obviously people are doing other things with coins on deposit, so the bad-debt metric is going to get replaced with something more closely related to the stated use of the coins, including how transparently that is disclosed.

[VEscudero]

I feel identification should be a highly weighted metric contributing to their final credit rating.

Without the supply of this identification they are just facelesss names on the net and i would suggest that people are capped at a A rating or lower without supplying this identification.

+1

[PatrickHarnett]

Some weeks ago I looked at an enhancement to the credit ratings I do because the range of businesses did not fit within the lend/borrow model it was initially designed for.  I have made changes to the routine and will be collecting/updating information as it becomes available.

• Possible size breaks would be 100/500/1000/2500/5000/10000.
• Possible customer number levels could at 10 and double 10/20/40/80/160

I have also made a couple of other minor changes to break points to better reflect the range of issuer data.

[wachtwoord]

Hashking has reduced his 8 week deposit to  1.25%  weekly

[Haplo]

Hmm.. I've been thinking about this quite a bit. A few tweaks I have in mind:

1) The ratio of "liquid reserves" only needs to match current accounts (ie accounts that can be withdrawn from at any time), and should be between 20-30%. The remainder should be in exclusively short-term assets.

2) For CDs (which have a more limited liquidity requirement), the more important metric is how well the asset maturities match the debt maturities. In other words, if a lender has 100BTC in 3 month CDs, they should have not more than 100BTC in loans/assets with 3 month maturities. Assuming they keep proper liquid reserves, it's less important whether the assets mature exactly when the CDs are due than whether liquid reserves are sufficient and kept replenished.

3) The amount of BTC the lender himself puts down (equity/deposits) should be counted separately, assuming the lender keeps a policy of insuring losses with their own money. This ratio should be ideally 100% or better.

4) Other security practices, like setting up dedicated receiving addresses with accounts and not allowing withdrawals to other addresses, should probably be added under the "non-financial" section. A significant fraction of major BTC losses are security related.

[PatrickHarnett]

Hmm.. I've been thinking about this quite a bit. A few tweaks I have in mind:

1) The ratio of "liquid reserves" only needs to match current accounts (ie accounts that can be withdrawn from at any time), and should be between 20-30%. The remainder should be in exclusively short-term assets.

2) For CDs (which have a more limited liquidity requirement), the more important metric is how well the asset maturities match the debt maturities. In other words, if a lender has 100BTC in 3 month CDs, they should have not more than 100BTC in loans/assets with 3 month maturities. Assuming they keep proper liquid reserves, it's less important whether the assets mature exactly when the CDs are due than whether liquid reserves are sufficient and kept replenished.

3) The amount of BTC the lender himself puts down (equity/deposits) should be counted separately, assuming the lender keeps a policy of insuring losses with their own money. This ratio should be ideally 100% or better.

4) Other security practices, like setting up dedicated receiving addresses with accounts and not allowing withdrawals to other addresses, should probably be added under the "non-financial" section. A significant fraction of major BTC losses are security related.

Some interesting points.  Not all of them are applicable to the people offering services.  Some comments/observations:

1: Having higher liquid reserves can be useful, but at a 20-30% level would leave small (by customer number) or poorly diversified operations exposed to contingent events.  Also, the short-term asset market is not as liquid as many people would like. 

2: Tracking maturities in a "point in time" system is not practical and some CD issuers manage this internally, but not visibly.  It is a cash-flow management issues and relates to solvency (paying debts as they fall due).

3: Surplus or net assets - covered.  Should it be greater than 100% is debatable - that goes to risk profile and the nature of the business.

4: If I wanted to look at security, I'd also look at the kind of websites and other things people do, but there are some clever thieves out there.  That is part of the due diligence anyone should do before investing.  That might also assume coins are sitting idle which is unlikely.

[Haplo]

Some interesting points.  Not all of them are applicable to the people offering services.  Some comments/observations:

1: Having higher liquid reserves can be useful, but at a 20-30% level would leave small (by customer number) or poorly diversified operations exposed to contingent events.  Also, the short-term asset market is not as liquid as many people would like.

Not sure what sort of event would be made worse by having extra reserve cash sitting around. Thinking about it, though, the 1-month term for "current debts" is probably a good enough approximation for "short term".

2: Tracking maturities in a "point in time" system is not practical and some CD issuers manage this internally, but not visibly.  It is a cash-flow management issues and relates to solvency (paying debts as they fall due).

Cash flow is a pretty fundamental part of credit analysis :\. It doesn't really need to be tracked exactly, just to make certain that loan/asset terms have similar (or greater) liquidity as the deposit terms.

3: Surplus or net assets - covered.  Should it be greater than 100% is debatable - that goes to risk profile and the nature of the business.

The rule of thumb for non-financial businesses is 1:5 (20% debt) debt/asset ratio or lower preferred, and 1:3 (33%) is running on bankruptcy. For a financial business, higher ratios are acceptable, but the considerations are different. Frankly, if a business chooses a higher risk model, they probably should not have as high of a credit score, and higher leverage means lower ability to repay principle in the event of some unexpected adversity. It does depend on the nature of the business, though, since a mutual fund would have very different considerations (and expected higher leverage) than a more money-market oriented "bank".

There's also incentive to consider, ie a mutual fund owner who charges clients even when he makes a loss does not deserve a high rating.

4: If I wanted to look at security, I'd also look at the kind of websites and other things people do, but there are some clever thieves out there.  That is part of the due diligence anyone should do before investing.  That might also assume coins are sitting idle which is unlikely.

I was thinking more along the lines of the recent forum hack whereby the hacker(s) gained access to other people's accounts and then made withdrawal request PMs to be delivered to a hostile address (client rather than server side compromise). On the other hand, security is probably beyond the scope of credit ratings anyway.

[PatrickHarnett]

Again - thank you for your points and thinking (and your posts in various other threads).

One of the things to keep in mind is that the people taking deposits are a wide mix of operating models.  Some are more like banks or financial institutions, others more like start-ups seeking capital for expansion or working capital to do something (maybe mining, maybe trading).

Fundamental to the ratings is if there are assets to cover liabilities, and if things go wrong, how badly wrong does it go.

If I had one customer with a 10000coin on-call deposit, then redeeming that could give me quite a lot of difficulty.  If it is 100 customers of 100 each, that  is much easier.  That is why people should ask to see what the various metrics are.

Also, when I am choosing an investment, I definitely ask questions, including deposit and withdrawal addresses.  For example, with Starfish I run a modestly sized address book with details on 200 customers (receiving and sending) and expect people I deposit with to be able to track my funds.  There are people I am happy to invest with and others not (and there are size breaks too) thus I deposited 2000 coins with someone yesterday, and another person just 10 coins.