BBY Disruptive Lunch: CBA's Kelly Bayer Rosmarin

It's sometimes easier to write about the "less" polished presentations we get at the BBY Disruptive Lunches then those that are focused and tested. Commonwealth Bank's (CBA) Kelly Bayer Rosmarin presented a cogent and polished summary of that institution's approach to disruptive technology generally and specifically to the digital client interface.

Bayer Rosmarin's engineering background was to the fore as she set limits on the subjects she was going to discuss and stuck to those parameters. Part of the charm of these lunches is the way that some of the start-ups meander through their business cases almost rethinking what they're saying as they go. I don't expect young companies to have laser sharp focus. When they do, as was the case with someone like Catapult Sports it can be riveting, but it's not always necessary. Take Pocket Book's (19  Feb 15) offering last month, that to me was notable because I got the sense the team was still discovering their capabilities and a best avenue of action.

Bayer Rosmarin sought to deal with disruptive technology and CBA through the lens of three distinct examples: Crypto Currencies, Cognitive Computing and Cyber Crime.

1. Crypto Currencies

Would crypto currencies exist without the 2008 financial crisis? Security, trust, convenience and efficiency, all areas severely tested by the crisis. The collapse of major banks severely tested the faith of people during the GFC. Historic examples from past economic adventures (think Weimar Germany, South American banana republics, African hyperinflation) provided fertile ground for Bitcoin (and other crypto currencies) in recent years.


Bayer Rosmarin's Bitcoin case study was a gentle debunking of the currency.

Firstly, coming from the angle of a large bank the security aspect of Bitcoin seemed to CBA somewhat counter-intuitive. Approximately 10% of Bitcoins would never be used because the security system functioned on the basis of a one-time passcode. Lose that, and you lose your ability to redeem your "money". Imagine (she suggested) if CBA froze your account permanently if you lose a passcode. The bank would likely to suffer an outflow of customers. How then could Bitcoin pass a convenience test?

Secondly there are a limited number of Bitcoins. I was somewhat confused by the CBA argument here because as we know Bitcoins can be split into fractions down to (I believe) nine decimal places. Additionally the raison d'etre for most Bitcoin believers is that the cap on the eventual size of the Bitcoin pool serves as a hedge against their concerns about central bank incompetency and associated currency debasement.

Thirdly, and more convincingly Bayer Rosmarin pointed to concerns as to the security of Bitcoin exchanges given various scandals that have erupted in the last two years. It's hard even for ardent Bitcoin enthusiasts to rebut this because the evidence is strong that this has been the weakest facet of the currency.

Lastly, Bitcoin according to CBA lacks differentiation. In fact, anyone can start a crypto currency. I think the suggest here was that these currencies might sink or swim as a group. If one currency was exposed as having certain vulnerabilities, then the group was likely to suffer because the flight to safety element of say out of Euro and into Swiss Francs wasn't available.

Not everything about crypto currencies is negative. Various protocols for payments were worth exploring. CBA have recently paid $A40 million for South African-based Take Your Money Everywhere. "TYME" is one of those bridging technologies that gives lower end or under-serviced users access to regulated bank accounts. The protocols and tech derived from this business is likely to add growth for CBA across not only Africa, but also in Asia. Bayer Rosmarin's analysis is clearly concluding that crypto-currencies are likely to stay as niche speculative ventures if banks can link their services in a way to make their day to day use unnecessary.

2. Cognitive Computing

The roots of today's Cognitive Computing can be traced back to the work done by IBM's lab in the mid 90's. That research culminated in the famous series of chess matches between supercomputer Deep Blue and Garry Kasparov.


Fast forward to today and think self-driving cars and algorithmic securities trading. The key to each step in these developments was the ability to handle ever-increasing amounts of data. "Big Data" as readers will know is an ever present theme at these lunches. For institutions such as a big bank, the question becomes one of sourcing and distributing access to that data, and what to charge or pay.

If you're a CBA SME customer, you can access some of that data through their Daily IQ application.


As a further add-on, CBA offers an open interface called Pi. The front-end of Pi is "Albert" and that gives you the opportunity to develop your own ways of interacting with clients.


Interestingly I allowed my mind to stray to some of the many presentations we saw last year which purported to provide just this type of add-on to users. As a friend of mine who runs a restaurant said when I showed him Posse (https://posse.com): "Why should I pay for that? I think my bank already offers the same thing?" What it says to me as an investor is that many of the businesses I look at are predicated on a single business plan. Attract a buyer such as CBA before they put their mind to developing their offering in your chosen sector. I know that sounds obvious. I've lost count of the number of times I've walked out of presentations thinking I've just seen the next unicorn only to change my mind after a cooling off period.

3. Cyber Crime

Readers will know I'm a long time fan of predictive algorithmics. We saw these methods first adopted by the Israelis in order to be able to deploy their limited security forces to the maximum effect. Financial institutions are already using these same techniques to track fraud and security infractions within their businesses. Cybercrime remains a huge challenge. Biometrics is often put forward as a solution, but Bayer Rosmarin doesn't see this as a standalone answer to tech-crime.

If you can find a business with a unique slant on cyber crime, I'd grip on tight. The very fact that there is no current industry standard solution suggests to me that this is an area that investors have a chance to make serious returns. Maybe part "Local Measure", part biometrics is what I'm looking for to fill the gap?

Summary

I view Kelly Bayer Rosmarin's presentation as a nice counterpoint to much of what we've been privileged to see and hear over the last year of BBY Disruptive Lunches. Call it the "Empire Strikes Back" if you will. It was at its core a well reasoned and somewhat sobering reminder that big institutions are well able to fill niches that smaller more nimble entrepreneurs are claiming as their own. One question during the Q&A best summed this up. Bayer Rosmarin was asked about the challenge being made to CBA's business by peer to peer lenders. While not dismissing the sector out of hand, she did suggest that operators (e.g. Society One) might find their businesses struggling if interest rates started rising. If you think about it, that makes sense. P2P's model is based on occupying the spread between lenders and borrowers. As rates rise that spread is likely to narrow and with it squeeze P2P business. Ultimately I got the sense that Bayer Rosmarin modus operandi is not to jump at shadows. Like all good engineers, it would seem she prefers to analyze first and act after carefully weighing up the possible responses and solutions.

Ciao!

CAIA / Deutsche Bank Volatility Breakfast. Check your ego, you're a passenger on this bus.

The hardest thing you can do as a professional in finance is to listen to others who are experts in your chosen field. It requires a certain modesty and the ability to close your mouth. For me, it was very strange to be sitting in the CAIA / DB Volatility breakfast listening to a panel discussion with David Dredge (Fortress Convex Strategies Group), Jerry Howarth (36 South), Michael Armitage (Milliman Financial Risk Management) and Michael Winchester (State Super, Australia). The panel was moderated by David Walter (PAAMCO).

For starters I worked briefly with Dredge at Artradis, in fact; he interviewed me, and I really enjoyed the process for once. Dredge (ex-BT) was on the team that developed the NDF (non-deliverable forward) market for currencies in Asia during the early 90's. That was a seminal product for emerging markets, and he told me some great stories about the early days of that market. The good thing about David is that he's willing to look at a lot of left field ideas. I remember getting a good hearing on some research I had done in the Australian mortgage insurance market that was initially dismissed by the team, but got some encouragement from David that was appreciated.

I wrote briefly about Jerry Howarth in my last blog, and so had a good idea of what he was likely to say. Michael Armitage  is new to me, and to be frank it took me a while to understand where he was coming from as I hadn't researched the firm before the event.

Michael Winchester was meant to represent the ideal investor for these funds. My guess is that he may or may not be invested in any or all them. It at least seems that he has done his due diligence on each of the funds. If I missed a disclaimer about this, I apologize, but I'd respectively suggest some disclosure. So be it.

Moderator David Walter (PAAMCO) is an investor with whom I'm familiar. I had the discomfort of once having to front him in his offices in Singapore to explain some returns that were somewhat unexpected. At the time, he was with ABN Asset Management, and they were being sold to PAAMCO. I only saw him one time after that during a Goldman Sachs conference in Tokyo, where he was kind enough to not dwell on the past too much.

Given that the breakfast theme was volatility and tail risk hedging, it was unsurprising that two of the three represented funds (Fortress and 36 South) focused on producing best cheapest convexity. Milliman as an offering was a different beast. Milliman state that they aim "to stabilize the volatility of an investment portfolio". In the context of this panel that effectively means, give us your portfolio, and we'll try and provide cheap hedges that will help protect or enhance your returns.

The discussion followed the usual path of each fund stating their raison d'etre. Dredge's opening statement was very familiar to me. Fortress Convex are going to go out into the market and try and buy the cheapest convexity. That means that you'll never know whether the hedge has a direct correlation to your portfolio as the purpose is to provide exceptional returns in extreme circumstances (3 standard deviation movements). If you're a long-only equity portfolio, you should get negatively correlated returns during "crash type" events. Fortress choose the best bets, whether that is (say) KRW/Euro or Canadian interest rates, and then that's what they'll own. So long as its positive convexity. The difference between Fortress and 36 South is that Fortress look for catalysts as part of their analysis. Howarth and 36 South are not concerned with specific events; rather they're looking at situations where volatility is cheap versus historical norms. It's a subtle difference, and I'm not sure the audience picked up on the fact. Both want to deliver a product that offers cheap convexity, but Fortress partly produce their "cheapness" by mitigating the cost through scenario analysis. 36 South do the same thing by a relative "cheapness"; the expectation being that the position itself has an innate value likely to offset some of the theta costs of holding the derivative position.

Milliman's direct hedge model is difficult for an independent consultant like me to assess. So much of what they do seems to rely on getting under the hood of a particular fund. How can I go to an investor and say that this company is likely to be successful in dampening the volatility in your portfolio and ipso facto improving your Sharpe Ratio if I can't do the math independently?

David Walter I felt hard a hard job getting Michael Winchester involved.  It seemed to me that the panel was leaning heavily in the direction of a consensus that volatility as a whole was cheap, and the dangers in markets were growing. Consensus is never as interesting as conflict and I would have liked Winchester to reflect on what his preferences were rather than to be in sync with the panel. He did allude to the difficulties he as a strategist confronted when trying to hedge volatility in a portfolio so concentrated on its exposure to the Australian banks. His conclusion seemed to be that historically as correlation went to one in disasters, then the basis risk was acceptable.

One other thing struck me as missing, and that was some discussion on the costs of the three offerings. I know 36 South is a conventional "2 & 20" fund, and I guess Fortress is similar. I have no idea what Milliman charges. Let's assume from the literature that the theta costs of 36 South and Fortress are in the region of 50bps per month. That means that if nothing happened in a year the likelihood is that you'd be down 8% (12 x 0.5% + 2%). That means that either you, as a fund manager have to produce enough excess alpha to pay for this, or you have to rely on the convexity provider to reduce this cost through clever positioning. I know with 36 South I get a reasonably long track record, and thus can weigh the propensity of the managers to achieve this. In the case of Fortress, the track record is shorter, and I'd have to add on to it David Dredge's previous track record to illicit the same conclusion.

After the breakfast, I was asked by an old associate what I thought. He liked 36 South and knowing him to be reasonably quantitative in his approach I could see the attraction. I'm somewhat conflicted on the argument that 36 South's model of outright cheap volatility is better than the Fortress catalyst enhanced model. If I'm 36 South, do I take profits too quickly if I didn't factor in a particular scenario? If I'm Fortress do I overplay my hand and get too stubborn regarding a situation not playing out, always increasing my bet in the hope it triggers?

Milliman is a different animal entirely. It seems I'm being asked to time the market more. Why would I take on their overlay if that's what the offering is if I'm bullish on the market? How much better is their offering than buying zero cost collars in index options? If a client asked me to investigate Milliman with them, I'd be very interested. It's hard not have an open mind about a company that has been around so long and is thriving.

After the main panel discussion had concluded, we got 30 minutes from DB derivatives strategist Alex Staab. A quick look at four  particular trades and an overview of what DB thinks is likely in the volatility markets. I thought it was particularly interesting that Staab pointed out the unusual correlation between the advancing Nikkei, QE and an upturn in volatility. The assumption in the "Anglo" markets has been that low rates, QE, and a rising market has dampened volatility. DB is suggesting that the Japanese pattern is repeating itself in Europe. I wonder if that means that Europe will have the same economic growth patterns as Japan had after its real estate market crashed in the early nineties. Let's hope not. Having said that It is true that Japanese volatility was extreme especially during the recapitalisation of the banking sector around the turn of the century.

Staab didn't get me with every point he made. I have to say a wry smile came to my face as he explained the positive return drift in daily v. weekly variance swaps. The trouble with that trade has been always to find someone to take the other side at near the theoretical value. I know at UBS we did our first trades in this in 1999 for the Japan book. What Staab didn't point out was that the seller over the swap makes a chunk of their margin in the hedge by trying to beat the end of day or week marks. On top of that there's a compliance problem because the bigger you are, the more likely it is that you can influence the close of the futures market. Regulators don't like that as it looks like market manipulation. This means that after a while compliance probably taps you on the shoulder and puts volume limits on trading around the close so as to avoid influencing things. My bet is that if I went to DB and said sell me the swap I'd pay a bigger premium than I wanted and be restricted in size. I'm fine with that, but others in the room who were more actuarial and less market experienced might be disappointed.

Overall I enjoyed the morning and will do some follow-up in the coming weeks. If you're interested in my conclusions and wish to discuss please contact me through the website at www.ibcyclist.com.

Ciao!