ECB View: Eurosystem prefers private sector bonds over govies – Purchasing data relating to the ECB’s QE programme over the month of January has confirmed our view that the ECB has indeed decided to decelerate more of its public sector bond purchases in favour of private sector bond buys. The monthly data is in line with earlier published data, which was on a weekly basis, and shows that the composition of purchases has been altered, while overall stimulus has been reduced. From the start of the year, the ECB has lowered its monthly purchase amount from EUR 60bn to EUR 30bn. We argued that while overall stimulus is reduced, the ECB would decide to focus its purchases more on private sector bonds, because of two reasons. Firstly, the ECB would avoid running into problems to find enough public sector bonds while still adhering to its self-imposed issue(r) limits. Secondly, a reduction of stimulus could lead to an unwarranted tightening of financial conditions. By focusing on covered and corporate bonds, the impact to the real economy would be mitigated. We expected that the absolute purchase amounts of the private sector programmes would remain unchanged. As a result, the private sector buys would increase as a percentage of overall purchases.180205-Global-Daily.pdf (50 KB)
Indeed, the data shows that the purchases of in particular covered bonds for January at EUR 3.3bn is very close to the EUR 3.1bn average from April to December 2017. The corporate bond purchases for January at EUR 5.6bn were only a fraction lower than the EUR 6.3bn average for April to December 2017. As a result, this means that the combined amount of both private sector markets (EUR 9.3bn) over the month January was very close to the average of April to December of last year (EUR 9.5bn). Since the overall monthly purchase amount has been lowered and the private sector buys have been kept stable, the proportion of private sector purchases has risen to 31% of overall buys. In the period April to December, the proportion of private sector buys was on average 16%, which means that the private sector buys have actually become twice as important in the ECB’s QE programme. In addition, we note that the deviation of purchases to the ECB’s capital key for the public sector programme (mostly government bonds) has declined. This means that the ECB has decelerated its overbuying of French and Italian bonds, while purchases of high scarcity countries like Germany and Finland have returned closer to the envisaged purchase amounts based on the capital key. Going forward, we expect that the ECB will have no difficulties in sourcing enough private sector bonds, given the room to the issue(r) limits and because of new supply. (Kim Liu, Joost Beaumont, Shanawas Bhimji)
Euro Money Markets: Just leave Eonia up to the ECB – On 1st February 2018 the European Money Markets Institute (EMMI) concluded as part of its Eonia Review Data Exercise, together with data published by the ECB, that pursuing a thorough Eonia review is no longer appropriate. This follows its earlier communication where EMMI published its ‘Consultative Paper on Enhancements to the EONIA Benchmark’ on 8th August 2016. The purpose of this review program was ‘to undertake any necessary enhancements to underpin the robustness of the benchmark’. Additionally, EMMI sought to align it with the requirements of the EU Benchmarks Regulation, which entered into full application on 1st January 2018. This entailed ensuring a reference index which is ‘robust, reliable and resilient and which can be calculated in the widest set of possible circumstances’.
In the second phase of the review, EMMI concluded firstly that interbank lending activity has a ‘fair degree of concentration’ that can easily skew the benchmark rate by a single contributor. The interbank activity also displayed a geographical concentration. Secondly the inclusion of non-eligible financial instruments doesn’t impact the representativeness of the benchmark. In line with the main conclusions of the ECB – that there’s more data available of borrowing transactions than for lending – EMMI indicates that banks are more active on the borrowing side than the lending side. The ECB also concluded back then that banks do not only rely on unsecured interbank funding as part of their total unsecured overnight borrowing. This is as transactions with other financial corporations (including those outside the Eurozone) represent a large portion of the unsecured borrowing.
Taking these factors into account EMMI has concluded that Eonia would need to be redefined as an overnight wholesale borrowing benchmark. This in turn would require using data of the borrowing side where volumes are higher and include borrowing transactions with a broader range of counterparties. Also, EMMI believes that as long as the definition and calculation methodology for EONIA remains as it is, then EMMI can’t guarantee a reliable benchmark rate and will keep the daily publication of Eonia as is. Finally, as we mentioned earlier – see here the ECB still plans to introduce a new overnight benchmark rate by 2020. (Fouad Mehadi)
US Macro: Non-manufacturing ISM points to stronger payrolls growth – The ISM non-manufacturing index rose to 59.9 in January, above consensus (56.5) and our forecast (58.0). This was a new post-crisis high, beating the previous recent peak in October (59.8), and the highest since August 2005. Notably, the employment index rose to the highest level since the series began in 1997, at 61.6. While not perfectly in sync, the index moves broadly in the same direction as services payrolls, and suggests we could see a substantial pickup in employment growth in the coming months. With the labour market already looking tight, this could mean additional upward pressure on wages, which have already shown signs of accelerating. Indeed, when the employment index was last near these levels in mid-2015 we had a number of months where payrolls growth exceeded 250k. Such strong employment growth would likely give the Fed greater confidence to continue rate hikes over the coming quarters. (Bill Diviney)