Always wanted to come back to this thread, but didn't have a reason to do so. Nway I'd like to revisit the issue of yield curve arbitrage covered in post 7 - 17 due to the recent developments in money markets.

This week we have seen market players favouring the reverse repo injections versus the traditional interbank mart suggesting some sort of imbalance in the distribution of liquidity. The CBK has been accepting lower and lower rates on the repo resulting in the negative spread in favour of repo (indeed this is unusual).

So in a situation where the interbank rate is at say 1.28% and reverse repo at 1.10% is it time for money market players to consider using the repo to exploit the yield curve arbitrage more effectively?

The interbank has been below 1.30% since September last year. There was a lot of government spending in June last year. The last time interbank was 5.00% was way back in May 2009 then CBK introduced reverse repos.

You need to ask yourself why reverse repo rate are lower than interbank?

From my rudimentary understanding, the reverse repo is emerging as one of the preferred mechanisms for central banks to mop up liquidity on short notice, in which case they may largely be trialing the effectiveness of the instrument? Believe the US Fed trialed this effectively last year. My suspicion is that Kenya, like other emerging and frontier markets is looking for ways to manage the overwhelming hot money inflows as overseas investors hunt for yield farther and farther afield...without direct market intervention.
Anyway, I defer to the intrepid 2 scoobies (especially since you're either treasurers or extremely dedicated personal investors).

Out of curiousity, why would it be appropriate to look at the spread between overnight financing (interbank) and term financing (t-bills) and view this as an arbitrage opportunity?

@Scoobs. Ideally if you take the interbank rate to be the equivalent to an unsecured loan then you could argue that a collateralized loan should be cheaper. But the fact that the bank’s line of credit is tapped out then dictates that making a loan to such an institution is inherently more risky (explains why horizontal repo rates are more punitive). So if you incentivise the use of cheap emergency funding then you make it more attractive for banks to arbitrage the yield curve.

An illustration would be the RR injections on 28-29 March 2010 and 28 Dec 2010. They happened to coincide with the value dates for the last two 15 year primary bonds; where the repo rate averaged 2.410% in March, but only 1.450% in December while 15 yr yields jumped to 10.923% from 9.980%. CBK is partially funding the higher spread but influencing market rates as liquidity could have been mitigated by government spending.

Or taking a more recent example the 182TB yielding 2.464% on 20th Dec 2010 when the reverse repo was at 1.460%, then take the 4th Jan repo transaction at 1.101% used to buy the 182TB at 2.675% last week; CBK is paying over 62% of the spread through the subsidy (nine months ago, only 50%). It would appear that this strategy would be flawed in a rising interest rate regime and that the existing banking cartel is oblivious to changes in cost of funding when setting borrowing rates.

If indeed CBK lowering T-bill rates influence lending rates then they’ll be hard pressed to explain why since TB yields floored in July 2010 the interest rate spread has risen to a five year high of 10.44% for four straight months. The spread was this high in 2004 but back then the fall in TB yields saw the interest rate spread fall from 14.81% in May 2003 to 10.33% in May 2004 (working old formula).

My solution is that whether through government spending or the reverse repo the money creation principle is the same. My theory is in order to maintain a heavily borrowing programme while taming the cartel, why not spend money in Treasury coffers (for liquidity) then mop it up through treasury auctions. Simply shifting money between M1 and M2, but exerting minimum influence on market rates; in fact historical precedent suggests that CBK could make it worthwhile for banks by paying interest on Treasury reserves held at CBK (regulated payments).

The deficiency in liquidity required for onward lending to ordinary Kenyans could be managed through the reverse repos at higher rates that would be motivated towards the squeezing of interest rate spreads. The last three weeks may have shown that banks have an affinity for the reverse repo as a signaling tool rather than the CBR; it’s a rate that CBK can use to exert influence.

I don’t think it’ll be possible to get rid of the banking cartel as long as government borrowing is crowding out the private sector. I’d like to think the problem here is government’s inability to spend money efficiently and quickly on it's projects.