Contact details:

[email protected]

Website:www.acarate.com

Invisible banking

Piyush Gupta, CEO at DBS was the keynote speaker at the most recent ACTS/ATC Forum in Singapore. DBS’ mantra of “joyful banking” has always irked me. I feel that banking is about execution of primary transactions that are important to me – I don’t feel happy about a mortgage, it’s the house that excites me; when I buy a product or service that I want, I just want the payment to go through without hassle.

So it was great to hear Piyush proclaiming that his goal for corporate banking is that it should be invisible – integrated seamlessly into business processes to reduce friction to the minimum that is secure. Invisible banking is a goal I can get excited about!

Treasury technology penetrating Asia at last

Asia has traditionally been a laggard in adopting treasury technology – penetration rates for SWIFT connectivity, TMS and eFX are generally lower than in Europe and the USA. So ACTS decided we should have debates on TMS vs XLS and eFX vs phone to stimulate discussion.

The debates were hilarious, even if the poor treasurers who ended up debating for XLS and phone were not true Luddites, they did a great job of getting into their roles.

And the results were very positive for the more technically inclined amongst us. On TMS, two thirds of treasurers were pro TMS. Good news for TMS vendors, and treasury efficiency in Asia going forward. Interestingly, TMS clearly have room to improve on the ease of use/flexibility front since half of TMS users still rely on XLS as well.

Similar data but different story on eFX. Ninety one percent of treasurers were pro eFX after the debate, but usage lags this enthusiasm.

Two thirds of treasurers use eFX, but half of those still need the phone as well. The reason is that, with many emerging markets in Asia, many Asian treasurers feel that they still need to deal EM FX the old way. India used to have a bank cartel (FEDAI) blocking eFX and many treasurers are not aware that RBI issued clear instructions encouraging eFX a few years ago. (There is a similar and equally wrong idea that SWIFT is not allowed in India; RBI also cleared this up a few years ago, and in fact the latest INR clearing systems use ISO 20022 messaging.)

I recently checked the status of eFX in Vietnam for a client and learned all three eFX platform providers I asked are active there. Honestly, I was surprised. It is a good indication that phone dealing is rapidly heading towards redundancy in Asia.

The limits of quantitative economics

I also had the pleasure of co-chairing EuroFinance in Barcelona. The conference’s first plenary speaker, Pippa Malmgren, had many challenging observations to share. Given her background as an economist, I found her warnings about the limitations of quantitative analysis bracing.

Paraphrasing, she suggested that we need to supplement quantitative analysis with something along the lines of ‘common sense’. When the numbers do not make sense, we trust them at our peril. She cited plenty of examples from the past decade to support her assertion.

I am a big fan of fact based decision making. Equally I have seen that many quantitative techniques we rely on in treasury have significant limitations. Hence, we supplement value-at-risk with stress tests, and hedge optimisation with common sense.

Conversely, I think it is very helpful to challenge gut feeling and common sense with hard data. So, I suppose my conclusion is we need holistic and open-minded decision making.

Will treasury be freed or terminated?

The second morning kicked off with an engaging discussion of the implications of AI and machine learning for treasurers. (The conference theme was “intelligent treasury”, and these days the word intelligent seems to be more associated with machines than with people.)

Long story short, the clear conclusions were first that AI will not be that intelligent for a while, even if its advances in specific domains such as facial identification and go are impressive, and second that treasurers will need to learn to work with AI more than risk being supplanted by it.

On why we still need treasurers, the blackbox issue came up – basically software is a blackbox and we cannot be sure what it is doing or why, so human sanity checking is a valid mitigant.

The clear upside is that simple improvements like robotic process automation will relieve us of the boring routine work and free us up for more interesting and higher value-added contributions to our companies.

Technology and what really matters

The rest of the second morning focussed on many of the buzzwords that seem so urgent. Big data can help us find interesting patterns in heterogenous data, potentially offering a way forward without the pain of restructuring our data.

Blockchain may advance industry wide collaboration to the benefit of all, even if its power may be more in an idea around which banks can collaborate rather than in the technology itself. (Banks are mainly working on permissioned (ie private closed loop) blockchains, which are functionally more like SWIFT or DTCC, rather than permission-less blockchains like bitcoin where no one is in charge.

AI will not be that intelligent for a while, even if its advances in specific domains such as facial identification and go are impressive, and second that treasurers will need to learn to work with AI more than risk being supplanted by it.

Refreshingly, Simon Taylor from 11:FS cut through the hype and stated that, for corporate treasurers, the first widely used blockchain is at least 18 months away and will probably be in the trade space.

A panel on fintech clearly confirmed that the term is practically meaningless – fintech is being done by banks, by large companies (ApplePay, AliPay et al), and by startups. To paraphrase Deng Xiao Ping: it does not matter who does it, so long as it works.

My big take away is that technology does not matter, solutions matter. We need to focus on clear articulations of pain points (jobs to be done, problems to be solved), and then let the experts come up with solutions. Wanting to trial a specific technology may seem sexy, but it is probably a recipe for higher cost and lower benefits.

Disrupting banks

Brett King kicked off the third day at EuroFinance. While getting miked up he showed me his digital bank, Moven, which is more like a slick and mobile first personal finance site than a bank. I especially liked that Moven uses APIs to connect with other banks (in the USA) to give a holistic view of personal cash.

Brett’s presentation focused on corporate banking, rather than Moven which is retail. One of his mantras is first principles design. To reduce the cost per kilo to orbit by 95%, SpaceX had to start with a clean slate. This is very hard for incumbents to do. Many bank CEOs want their organisations to disrupt themselves, but the operational realities of staying compliant and maintaining trust are antagonistic to disruption.

A good illustration is SWIFT’s gpi, which will undoubtedly improve the classic four corner correspondent banking approach to payments – if only by enforcing decent existing practice on participating banks. It took an outsider – Ripple – to completely re-imagine international payments.

In China, where tech companies have been particularly aggressive in finance, banks are already out of the game in many spaces. Alipay’s YueBao is the world’s largest money market fund. One bank CEO said, “We have already lost payments (to the tech companies), so what can we save now?”

The tension between the compliant and trusted status quo and the need for disruption to make significant improvements will be an interesting space to watch over the next few years.

Risk management learnings

The risk stream at EuroFinance provided some great case studies and panel discussions. The first afternoon revolved around FX hedging.

This was ably kicked off with a beautifully clear explanation from Vincent Delort, Risk and Reporting Manager at JTI of how to do cash flow at risk (CF&R) in Excel – down to the actual Excel formulae. In case you are wondering whether this was too technical for most conference audiences, the room was packed and we barely managed to get through one tenth of the questions during Q&A.

One question we did get to was why use Excel for CF@R rather than a TMS. The answer was that none of the TMS CF@R solutions is sufficiently flexible to meet JTI’s needs. Another opportunity for the TMS vendors to up their game.

The views and opinions expressed in this article are those of the authors