OIS to Predict Central Bank Action

shuvabrata · May 20th, 2009, 3:50 am

How do we find out the probability of a central bank move (rate cut etc) from OIS rates of that currency? Any good research paper / idea etc on the same?regards,Shuvabrata

MCarreira · May 20th, 2009, 10:58 am

Do you mean something like FFIP on Bloomberg ? There's a paper describing that methodology in there.

shuvabrata · May 20th, 2009, 11:51 am

Where is the paper? Can you kindly post it here?

MCarreira · May 20th, 2009, 12:08 pm

Try this link

freddiemac · May 20th, 2009, 6:35 pm

Hi! As MCarreira said the usual way is to look at options or futures on (central bank) rates. OIS could be used in the same way. An OIS is just average of the overnight rate of a period. The problem is when the o/n rate and the policy rate diverges. Then the OIS becomes a difficult tool to predict central bank rate moves. Look at EONIA-refi, intended vs effective fed funds or SONIA vs BoE rate... If you have a lot of volatility in the policy rate - o/n rate spread then the average o/n rate during a period can be quite different from the average policy rate. To model that under the current circumstances is difficult. I have not seen any papers that deal with this. If you however just assume that the policy rate and the effective o/n rate is the same then it is quite straight forward.. (and wrong)... Sorry not much of help.. This paper discusses the idea of predicting central bank actions but assumes that the o/n rate = policy rate. http://www.rbnz.govt.nz/research/bullet ... hoy.pdfHTH

yoshimura1 · May 21st, 2009, 12:36 am

interesting. any more papers discussing ways to extract market expectations of central banks' actions?

shuvabrata · May 21st, 2009, 7:59 am

QuoteOriginally posted by: freddiemacHi! As MCarreira said the usual way is to look at options or futures on (central bank) rates. OIS could be used in the same way. An OIS is just average of the overnight rate of a period. The problem is when the o/n rate and the policy rate diverges. Then the OIS becomes a difficult tool to predict central bank rate moves. Look at EONIA-refi, intended vs effective fed funds or SONIA vs BoE rate... If you have a lot of volatility in the policy rate - o/n rate spread then the average o/n rate during a period can be quite different from the average policy rate. To model that under the current circumstances is difficult. I have not seen any papers that deal with this. If you however just assume that the policy rate and the effective o/n rate is the same then it is quite straight forward.. (and wrong)... Sorry not much of help.. This paper discusses the idea of predicting central bank actions but assumes that the o/n rate = policy rate. http://www.rbnz.govt.nz/research/bullet ... pdfHTHThis paper assumes that cause for Aussie and Kiwi the o/n rate and the policy rate are actually the same.... let me go through the paper and then i can comment better on the sameregards,Shuvabrata

freddiemac · May 21st, 2009, 8:34 am

QuoteThis paper assumes that cause for Aussie and Kiwi the o/n rate and the policy rate are actually the same.... No they are actually not the same. See eghttp://www.rba.gov.au/MonetaryPolicy/about_mon ... _policyThe reason is simple. The OIS mimics a strategy that places cash in the interbank market every day. Often the policy rate and the interbank rate is closely related but they do not have to be. Since you cannot place cash at the central bank at the policy rate divergence can occur. HTH

cpulman · May 21st, 2009, 9:09 am

Hi,I wrote the attached paper a couple of years a go. In the light of the financial crisis it's out of date (and certainly has a few errors in it) with respect to the FRA/OIS spreads and how EONIA etc have traded since August 2007, but it should give you a bit of an idea about how to go about doing this practically. Due to the market requiring my constant attention I never had time to update it, I'm afraid, but may do in the future.Happy to attempt to answer any questions you might have, but am not able to trade this stuff at my new place so I'm a bit rusty...Cheers,cpulman

Panang · May 22nd, 2009, 10:12 am

Agree with the points on eur/usd/gbp ois vs base rate. AUD and Kiwi ois swaps actually do fixed to the base rate, there is no basis. You can find this on the ISDA website I guess.

Martinghoul · May 26th, 2009, 11:00 am

QuoteOriginally posted by: yoshimura1interesting. any more papers discussing ways to extract market expectations of central banks' actions?This one is my favorite: Federal Funds Rate Probabilities