Cost of Capital

[HitmanH]

I'm trying to do a study of the cost of capital for a handful of desks / businesses - some internal to a bank, and some with hedge funds / prop shops.



I'm starting with some rather simple assumptions:

A) Within a bank - be it a prop business or a customer facilitation business that holds inventory overnight - they charge for that

B) within a hedge fund / well capitalised prop shop - if one has say 10m - you only pay for balance sheet as of 10m USD +1USD - on that 1st USD

(please let know if anyone does not agree with this)



And the question:

1) Is the cost the prop desk's / internal desk's cost of capital i) their cost of capital; /or internalised benefits

2) Do banks internalise / utilise customer balances / to reduce their cost of capital?

3) Different prop desks - from a HF / Liq Prov desk vs a super-levered EM/special sits desk - do they get the same costs?

4) If they don't do #2 - do banks or broker/dealers - on a whole - finance out assets as per their own tenor - or typically (as I believe) finance their own prop to a set period (90d)



Ultimately, I'm trying to work out the costs of capital of internal prop desks of banks, from the GSPS / special sits style businesses, to the liquidity provision/HFT strategies with v.high Sharpes.



Maybe I'm guessing here - but people like Nounis - who evaluates people who have gone from Prop > Fund - hoping you have some date points. Along with ANYONE else please - any help much appreciated.

[Praetorian]

I can give you some color about a case I have seen (I don't know if it is representative and it certainly was no trading venue). The Bank had three funding sources: Loans/Bonds, Customer deposits and ABS funding. The cost of capital was IR + credit spread for all units of the bank. Each cash flow within the bank was splitted into margin result, interest result and liquidity result. Depending on the business function they then calculated their result. Margin result of deposits (Cost-of-Capital minus IR) was credited to the Retail function. The funding benefit of ABS was credited to the ABS people (Cost-of-Capital minus ABS rate). Liquidity and maturity transformation was credited to treasury. Assumption B did not apply, the capital of the bank was assumed to carry some interest, otherwise treasury would have benefited from that.



The idea behind that is that each result is generated at some business unit and you want to measure each business unit alone. This idea is pretty simple. Unfortunately it gets very messy if you have international business with funding in several currencies, lots of subsidaries etc. To work the details out, it took 40 people over 3 years...

[gw33]

have looked at this in the past - would be very interesting to discuss. A lot of the prop desks will wrap it up into a broader number as the cost of seat in my experience.