CBSL two-way quotes: Realistic or elusive?

By Mahesh Senanayake 

It is a well-known factor in the in the primary dealer industry that the CBSL’s two-way quotes, commonly known as published yields/rates with regards to selling and buying bonds are not the rates primary dealers bid in the auction. Therefore, it can be argued that the CBSL published rates and the real market rates are always different to each other. 

A careful study on the said matter will reveal ample number of evidence to support the argument. Hence it is worthwhile to observe the practices related to how CBSL arrive at relevant rates, how the market rates are determined and how disparity occurs. In the market arena it is the buying rate that is considered far significant than the selling rate. Buying rate is the rate at which a primary dealer would by a security and the selling rate is the rate at which a security is sold at the secondary market.

Basis for CBSL two-way quoted rates 

The industry experts as well as the CBSL officials with years of experience are well aware that the CBSL two-way quotes (bid and offer rates) are determined based on the information obtained from the primary dealers who usually submit the rates as a daily routine based on their market transactions and previous records. Further the rates determined at an auction are very much unlikely to match the CBSL published rates. It could be either higher or lower than the CBSL published rates. It means the CBSL rates, market rates and the auction rates usually indicate different figures. 

Thus once the CBSL collects these details from all registered primary dealers, an average of the rates indicated by the primary dealers on each maturity is taken as the CBSL published rate. This means there is an absence of scientific method to determine the real prevalent market rates except the data generated by the Bloomberg trading platform or reports generated by contacting money brokers.

Likewise there is no guarantee that all primary dealers update the CBSL their true buying and selling price due to certain policies and administrative issues. For example, a certain expert pointed that lack of communication between the front and back offices of the treasury departments in some primary dealers has led to this situation and this was pointed out several times to CBSL.

It is widely accepted the rates which are transacted in the Bloomberg trading platform are the actual prevalent market rates which reflect the actual rates determined by the demand and supply (buying and selling). The wide acceptance is based on the fact that it is a must for all primary dealers in Sri Lanka to trade on Bloomberg platform or report all the deals which are done through the money brokers. This means when a primary dealer carries out a transaction on the said platform, the rates accepted will automatically be captured on the Bloomberg platform.

Further in the event the transaction is done through a money broker, no primary dealer has the luxury to conceal the transaction or the rates as per CBSL regulations and they are bound to report the details of the transaction to Bloomberg platform within half an hour. Therefore one can argue that the Bloomberg rates are more realistic than the CBSL published rates.

Some examples 

of the disparity 

Several example occasions can be observed to support the argument that the CBSL published rates, market rates and auction rates are different to each other. For example on 27 December 2016 CBSL conducted a bond auction calling bids for three maturities (1.3.2021, 1.08.2024 and 1.8.2026) offering Rs. 19 billion each. The results of this auction are published in the CBSL website (http://www.cbsl.gov.lk/pics_n_docs/02_prs/_docs/press/press_20161227e.pdf).

The CBSL two-way quoted rates which were relevant to the said auction is published on the same website (page 13 http://www.cbsl.gov.lk/pics_n_docs/_cei/_docs/ei/wei_20161223.pdf). (It is considered “relevant” as CBSL has continued to indicate the two-way quotes as the reflection of the market. And so far no professional dialogue has been ensued between the CBSL and the expert to dispute or agree with the said point.)

As per the 27 December 2016 auction, the weighted average yields are way below the rates indicated by the CBSL. When the four year and two months maturity is auctioned at 11.94 % the CBSL published buying yield is 12.15. When the seven year and seven months maturity was averaged at 11.98 at the auction the CBSL reported rate stood at 12.55. The nine years and seven months maturity was auctioned at an average of 12.11 when the CBSL published rate was 12.57. This example shows the disparity between the auction yield and the CBSL two-way quotes/published yields.

Several similar examples of this nature can be found, such as the outcomes of the auctions held on 13 January 2011, 10 February 2011, 24 February 2011, 30 January 2011, 28 February 2012, 10 January 2014 (of which details are available on the CBSL website) and occasions where the auction yields and the CBSL yields matched each other are almost non-existent, expect for a very few times. 

What are the probable conclusions that can be arrived at by looking at this kind of scenarios? Firstly, it is obvious that there is a disparity between the auction yields and the CBSL projected yields. Secondly, why would a primary dealer want to bid way below the CBSL rates when he has the opportunity to buy at a higher rate in the market? In the given example all auction based yields are way below than the CBSL published rates. 

Thirdly, how and why do the market players want to acquire Government securities below the projected yields? On one hand, it can be argued due to the competition created through the auction system, the bidders bid at the most competitive rates. On the other hand, a strong argument can be put forth that the captive sources have been used to control the rates especially when a huge disparity around 20 to 60 basis points is evident.

All in all, the above status quo is very well known among the bond market participants. What is unsavoury is that some commentators who elaborate on the Government security market use the CBSL published rates as the benchmark and calculate losses or profits, ignoring the actual outcome of the market activities. This is either a deliberate oversight or an effort to present a de facto scenario misleading the decision makers and the public. It is high time a national level discourse is pursued in order to arrive at a more pragmatic solution to avoid the disparity between the CBSL published rates and the auction yields.