From the desk of John Arnesen
Consulting Lead, Pierpoint Financial
2020 was, whether you liked it or not, a year full of data. Once the pandemic took hold, the Government rolled out data from rates of regional infection to the National Health Service capacity on a daily basis in March, and like many, I was thirsty for information and watched the 5pm updates religiously. I then discovered the Johns Hopkins Coronarvirus Resource Center website, which has to be the most comprehensive data set by country produced by any institution. Having been glued to it since mid-March, it was with a mixture of dread that I observed western countries losing any control over the spread of the virus to rooting hard for India that had kept it mostly out of the population, certainly during March. India did succumb to the rapid spread of the virus, but if you look at both the infection numbers and death rates, they have managed a distribution in a curve that should be the envy of other countries.
Somehow, India, with its one billion population and concentration of urban dwellers, has stifled the virus's spread. On September 16th 2020 they had 98,000 new cases and 1,290 deaths. As of January 10th 2021, these figures are 16,000 and 161 respectively. Contrast this with the USA that depressingly looks like it is still in the first wave and the United Kingdom that clearly shows how the UK mutation has reset us back to the level of the first wave of infections.
The data contains much information, and I found it a real help in managing my expectations. In early March 2020, I still thought I was going to Menorca for an Easter break. By December 2020, I expected my passport to gather dust for at least another year. The data tells me that globally, we are still essentially in the first wave, and even if we roll out the vaccine aggressively, other countries will need to do the same if travel is to resume. That will take time, and it appears countries are already suffering from logistical issues. So, vigilance and patience are the order of the day, and if any positive news emerges, it will show up in the data.
Securities Finance
Securities finance produces significant revenue year-after-year for principal participants, and it has mostly achieved this in an over-the-counter environment without the benefit of reference data. Asset managers have every pricing tool imaginable at their disposal to trade foreign exchange, cash equities, derivatives and ETFs but ask them where they should lend 270,000 shares of Deutsche Telekom AG (for example), and you’ll likely get a blank stare or a response along the lines of 'I outsource that activity to my agent lender' That all sounds perfectly reasonable, and the agent is given a fair amount of discretion in determining the fee it is willing to accept for lending the shares but what references does the agent have to know the value?
Ideally, an event has put Deutsche Telekom in play, and a decent number of borrowers are looking for the stock. That gives the agent price discovery, but that isn't the entire story. What is the collateral the borrower wants to post? Once you step outside of cash collateral which, by definition is the best common denominator, you have to assess the quality of the collateral, its liquidity, limits set by regulation, the clients or an internal policy at the agent lender. Moreover, do you even know what the collateral will be when you lend the 270,000 shares of Deutsche Telekom AG? If not, how can you price the loan? Surely there is a vast difference in value between receiving the Gilt, 0.5% 22/7/2022 over the corporate issue, Fnac Darty 2.625% 30/5/2026? In reality, when lending against non-cash collateral, the collateral is rarely known at the point of trade, and while the argument that 'one assumes the lowest acceptable quality of collateral 'prevails when pricing a loan in such circumstances, I don't buy it, I never have.
Fixed income lending is equally absent from price discovery. Lending against cash at least gives an interest-rate reference to reverse-repo of euro government debt on which to base the spread but yet again, that bid for cash will differ from borrower to borrower if they are willing to quote at all. They can trade repo all day long over Brokertec and take advantage of the liquidity it offers. As an agent lender cannot it is left to its own devices to price a request for a bond it hasn't lent in a while, if ever. This is further complicated by the margin/haircut differentials between repo and securities lending.
This conundrum was the driving force behind the rise of data providers which, to their credit have added and improved on the quality of data over the years and is no doubt beneficial to agent lenders in providing some transparency in where its peers have set fees on individual loans on an historical basis. It doesn't help you price a request right now in a live environment but acts as best as a reference to yesterdays aggregated fees. While participants can receive live price data, limited transaction volume, and aggregated data, in general, can pose challenges too.
Some years ago, we had a client with very liberal collateral acceptability guidelines and not wanting this to go to waste we found a borrower that was long a fair amount of said collateral and was keen to finance it. We lent them core European government debt (German, French, Dutch) on a term basis in exchange. Such was the size of our lendable government bond supply that we achieved this in two issues. The fee on these loans had nothing to do with the intrinsic value of the bonds but everything to do with the financing of hard-to-move collateral. Those loan fees showed up in the data providers’ output but was that indicative of where those two bonds traded? Absolutely not and if we were doing this, I think it safe to assume others were too. A similar disparity exists for collateral transformation trades. So, while the data tells you something useful about the issue, it is not the entire story, particularly when it comes to fees.
I used to work with a custody and securities lending salesperson who would always conclude the completion of an RFP with the following, 'here is what bad looks like, John …' He would then describe all the points in our submission that could be poorly received, where we would likely score less well than our competitors, human factors against us etc., etc. Once this diatribe was over, he would follow up with 'let me tell you what good looks like' and then, with enthusiasm describe how well-positioned we were, how our securities lending fee split proposal was more innovative than the rest of the market, how much the client liked the organisation and how we were going to win!
When it comes to lending securities here is a list that is by no means exhaustive of what 'good data' looks like for me.
A view of every loan in the same security with the fee, settlement date and duration by borrower.
A view of the percentage of my holdings compared to the lendable supply in the market, and if a corporate bond, the issuance size.
Details of the average length of loans per borrower, perhaps split by fixed, equity and country.
A screen that gives me a clear understanding of failed trades where the borrower failed to receive the loan due to their error and recall fails that is the borrower’s fault and the associated expense.
A snapshot of the volume of locates from a given borrower over selected period of time, the potential to fill the locate and the actual hit rates expressed as percentages. A more granular view of the reasons why a trade wasn't executed would also be useful.
Naturally, I want to capture all locates for the same security in an automated fashion and displayed instantly on-screen from other potential borrowers, favouring those that show a bid.
As an agent lender with such data at my disposal, one would be in a far better position to make an informed decision about the fee required, despite not necessarily knowing anything about the collateral if its non-cash which is likely in today's environment. Of course, you may be turned down, but at least you have a sense of what to expect from potential borrowers and after a year like 2020, (see chart below) which saw a continuation of reducing revenue that started in 2018, is it time to think more about extracting more value from the quality of loans you make rather than the volume? Data is invaluable, and anyone incorporating the list of features described above in their lending activity is a step ahead of the rest. The data exists, whether it is organising it into a digestible format is the challenge. Additonal demands to incorporate ESG data will undoubtedly evolve this year, so perhaps to kick off 2021 a review of data and future needs would be time and resources well spent.
