Monday, 9 March 2015

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


No comments:

Post a Comment