Roundtable: Out with the old asset allocation models and in with the new

Participants:

1. Pascal Blanqué
Chief Investment Officer, Amundi 

2. Didier Duret
Chief Investment Officer, ABN Amro Private Banking 

3. Daniel Kuehne, 
Head of UK Investment Solutions Group, Julius Baer

4. Bill O’Neill
Chief Investment Officer, UBS Wealth Management

5. Amin Rajan
Chief Executive, Create Research  

6. Jan Straatman
Global Chief Investment Officer, Lombard Odier Investment Managment

 7. Eric Verleyen
Chief Investment Officer, Société Générale Private Banking

Moderator: Yuri Bender
Editor-in-chief, PWM 

Yuri Bender: Welcome to FT headquarters for our roundtable discussion on the future of portfolio construction and asset allocation. Our aim is to discuss the inherent problems of the asset management industry, to discover new ways of working and to at least begin to map the shape of the industry’s future.

There is something in the human psyche, is there not, that leads us to draw up a theory for everything? Was there something inherently attractive about a ‘modern portfolio theory’ which propagated taking more risk for more reward, and led to several Nobel Prize-winning submissions and helped some consultants make millions in selling their computerised risk management systems?

Pascal Blanqué: We learned at school that the higher the risk, the higher the return, but in the last 30 years, low-volatility stocks outperformed high-quality stocks. So, to some extent, this is an enigma, and it challenges the theory. 

The crisis acted as a wake-up call, collectively. This has not just come out of the blue. We knew there was some misunderstanding or misinterpretation of some elements of the theory. Think of portfolio theory pioneer Harry Markowitz’ framework, where all individuals are supposed to have the same kind of expectations; so there were flawed aspects of the theory. 

Diversification, as we defined it in the last 20 years, didn’t work when most needed. We comfortably spent 30 years trading equities vs. bonds and called it allocation, and it was comfortable because government bonds provided the cushion for exposure to risky assets, but the efficient frontier of government bonds has changed, meaning their utility function must be reconsidered. 

And this opened the field for research, moving forward. We also failed to take full account of liquidity as the third critical dimension on top of risk and return for managing money.

Yuri Bender: Professor Amin Rajan, you were using this form of modelling back in the 1970s in the UK civil service. Was it problematic even then to model the macro-economic equivalent of modern portfolio theory? What conclusions did this lead you to reach about investing?

Amin Rajan: In fact, we were forced to model what was then called the rational expectations idea, which came out of the University of Chicago and had a lot of influence on the IMF, which leant the UK about £3bn (€4.15bn) in those days. 

However, when you try and model it, what you end up with is an equation, which, in plain English, basically means the future is not going to be very different from the past. In other words, the past is a perfect guide to the future, and you know and I know that in the world of investing, the past may be a good guide but it’s a very imperfect guide to the future. 

So when I left the Treasury, what I did was put some models into action to test how those equations were behaving. In almost every case where we had made predictions based on rational expectations, our forecasts were wrong. On the rare occasion when they were right, they were right for the wrong reasons, in the sense that the underlying assumptions we had made about exchange rates or the world price of oil turned out to be faulty but they helped us on this particular occasion. 

So from then on, I became very sceptical about rational expectations, and when I had a chance to write about this in the investment context, the amount of hate mail I received was breathtaking. I couldn’t believe how many people, particularly in the United States, really still believe in modern portfolio theory. 

If you look at the empirical evidence, and I surveyed 140 of the most prominent articles on modern portfolio theory, we found there was no evidence to support it.

And if you look at what has been happening in the field of investment since 2000, everything in that field suggests modern portfolio theory is completely irrelevant to today’s needs. 

Didier Duret: I think we all know what went wrong with modern portfolio theory, but we still have to come to terms with what went wrong with us, meaning: how were we able to really take it and fit in the notion of risk in one number? 

But I’m very happy to see that the asset management world is coming back to the idea that risk is asymmetrical, because, after all, this is the first assumption of private banking for centuries.

Bill O’Neill: Basically, the clients want guidance from us as wealth managers. They want to know how to allocate their wealth over the longer to medium term. For us, by the way, that’s five to seven years. And I’m talking here about strategic asset allocation. 

And at the end of the day, these big strategic calls account for about 80 per cent of the overall experience with a client, with TAA [tactical asset allocation] and stock selection at about 10 per cent each on either side. 

What we’re trying to do here is provide clients with guardrails in terms of their asset allocations over time. When we look at the world, regardless of what you feel pre-crisis or post-crisis, you still need a systematic approach to how to look at asset markers, individually and collectively. 

I don’t buy the argument that the crisis killed diversification. You just have to hang in there and hold your nerve, to keep focused on the medium- to long-term, and at the end of the day, a balanced portfolio in US dollars should present a return of about 6 per cent with volatility of about 8 per cent.

Jan Straatman: Before we take away the Nobel Prize from Harry Markowitz and throw away the modern portfolio theory, we have to understand that what he introduced was not just a model. It was much more a concept that was developed in a period when investors were investing only in cash and fixed income and only domestically.

If there was no global co-ordination in terms of economic and monetary policy, you automatically have low correlation globally. So what he introduced at that point in time made a lot of sense, conceptually, to create, a more global portfolio and investing in other building blocks that were low correlated, and you have a much better diversified portfolio. Those concepts still make a lot of sense. 

But correlation over the last 20, 30, 35 years has increased enormously due to globalisation, the introduction of large-scale derivatives which allow us to move quickly in and out and the role of very big financial institutions which hedge and arbitrage every minimal inefficiency that exists among traditional asset classes. 

Yet our entire business has been built on the modern portfolio theory concept. 

I’ve worked for a lot of institutional clients and have done a lot of ALM [asset-liability management] studies. I’ve learned painfully how wrong you usually are after three years with all the assumptions you put in.

There are very important flaws that exist, not so much relating to the concepts introduced by Mr Markowitz but because our world has changed dramatically. So it’s more the assumptions and  implementation of those concepts that is the problem, and the fact we have been quite complacent over the last 40, 50, 60 years, in terms of how we implement models. 

If curiosity killed the cat, then complacency will kill the portfolio manager. We still assume the monetary authorities are right, we still assume those models are right, and quite blindly follow the output. And that is the biggest problem.

Yuri Bender: Is it fair to say that investors have suffered from a lost decade in asset management, because institutions have failed to update and revise their models?

Eric Verleyen: I think maybe the model was not prepared to see the crash of the US market of 57 per cent during the peak of 2008 to 2009, and then we had all those portfolios of our clients suffering some huge loss. They were not prepared for this, because they were in balanced or even conservative strategies. 

Having said that, I wouldn’t say we have actually lost this decade, because some long-term investors were able to recoup their losses. Of course the recession may not have been good, but the market has recovered. So what is important is to make sure you don’t suddenly change your strategy. If you need to adjust the model, you should adjust it progressively, and that’s exactly what we are doing. 

More and more, we are looking at other sources of alpha, we are looking at factors impacting our portfolios. We’re not looking only at the correlation we used to look at and the matrix of covariance, which is exactly the modern portfolio theory assumption.

Daniel Kuehne: If we were sticking to equities and bonds, then variance would still be optimal. I strongly believe the world has evolved and we have many other asset classes, which need to be introduced and actually lead to the asymmetric distribution of portfolios. That’s why we need a new concept. Today’s opportunities make it much more challenging to model these aspects we have mentioned. We need to manage the expectation of our client. It’s most painful if they give up their strategy. That is fatal. We need to explain to them what to expect and what they need to prepare for in their portfolios.

Pascal  Blanqué: There are times in the cycle where you need to be concentrated, not diversified; when a single factor works for you, just like during the crisis. Concentration is probably more powerful than classic diversification. 

Secondly, I believe factor allocation is much more effective than asset allocation today. The classic approach in the last 30 years was to say asset classes are driven by macro-financial variables, and those variables can be predicted. We spent a reasonable amount of resources, time and money trying to predict macro variables. 

The new approach is to say risk factors are driven and polarised by macro-financial variables for which asset classes can be seen as proxies or agents. This paved the way for a different way of thinking and a different kind of organisation of the investment space. 

Didier Duret: Diversification is important, but around the table I think there are two camps. There are people who think strategic asset allocation as we’re used to working with it still works, and we have people who think we need to do things differently. That’s probably something we need to explore now, because it means a different way of managing money. 

We can still hold our nerve, as you say, meaning we stick to our guns and the bond vs. equity allocation will continue to work, or we go and explore some kind of proxies for bonds, because we know bonds have been totally disrupted by the regulation and central bank action and so on. 

Should we go with alternatives? Should we go into non-liquid assets? That’s the debate. But I think that’s probably the hedge that each investor has to decide: Do you stick to your guns with the strategic asset allocation, or do you want to explore new ways to compose strategic asset allocation?

Yuri Bender: Do you feel central bankers with their aggressive easing and post-crisis regulatory zeal have been the assassins that have fired that final bullet into the still-wriggling corpse of modern portfolio theory? 

Didier Duret: They have hammered the nails into the coffin by pushing down the yield to levels where the notion of income is called into question by any asset managers, pushing our work to its limit, where we need to question, where are the sources of income? Where are the asset classes we can diversify into? 

The role of bonds is questioned now. We need to find alternatives for bonds in terms of income and diversifying roles. The future is about looking at alternative investments, specific managers, sources of alpha.

There are no borders. If you look at the equity world, we are saying the correlation has increased, but in fact the world of equities has been expanding, the world of emerging markets has grown to some form of maturity. We have the frontier market, we have all the illiquid assets that we can have access to. I don’t see a limitation there in the field, in the scope for diversification. And this is really our job.

Yuri Bender: If we’re talking about this diversification, how will the choice for most wealth managers work? Will they rethink their strategic asset allocation and work with much more limited tactical bets such as thematic investment? 

Bill O’Neill: Thematic investment is an accompaniment to an investment process, or an output of an investment process. Discretionary managers would argue, for instance, that a lot of thematic investment is captured in their positioning. 

And I think it’s extremely important, but in terms of the way we’re looking at strategic asset allocation at the moment, it’s things like, for instance, looking at credit as a substitute for equity, lower returns relative to equities, but certainly volatility that is also lower. 

But in bringing it all together, you are still focusing very much on drawdown risk. Clients want to see worst-case scenarios, how well or how badly are things going to actually be in such situations? And that’s a concept clients can focus on, because essentially, it’s in pounds and pence.

Yuri Bender: Talking about pounds and pence, I’d like to flip the coin and look at it slightly differently. The option of old-style strategic asset allocation, combined with fairly timid, limited, tactical manoeuvres: does this still have a future, or do we have to be braver?

Eric Verleyen: We need to have this kind of long-term asset allocation, or at least your strategic asset allocation, and then to be able to adjust this. Having said that, what we’ve been saying about the change in the financial scenery, I think that you shouldn’t only rely on your strategic asset allocation and your classical-type bond investments. 

Going forward, if we look now at the output of a traditional optimiser, built in a decade where interest rates were going down, the optimiser may tell you that if you only base your decision on the past, you should continue to invest in bonds. 

Well, we all know that there is a big change, and that at some point, interest rates may go up. Of course, it may not happen this year, but at some point, it will happen. So only basing your decision on this when you build your portfolio wouldn’t be appropriate. 

How do you do it now? Well, maybe you try to diversify your strategic asset allocation. Maybe you could look at infrastructure, such as products based on the return or the income from roads, or maybe also social infrastructure – hospitals, prisons, this kind of thing – that would deliver you some income. 

You could also look at real estate. You need to look at other ways to increase your income. We are talking not only about tactical asset allocation, but completely changing the way you are thinking about your strategic asset allocation. 

Jan Straatman: A lot of the things that we talk about are basically around three elements. In order to solve a number of issues, you need to create more innovation on everything we’re doing. What are the building blocks I use to build my portfolio? What is the return stream they represent, and does that return stream make sense, from an objective perspective and from a correlation perspective? 

That may be traditional asset classes at some point in time, but in many cases it will be different types of things. In most cases, a more factor-based approach to building equity portfolios may be better, for the simple reason there aren’t any derivatives yet, there aren’t any indexes yet for these types of things. So that is a way for innovating your portfolio. 

The second thing is, how do you do portfolio construction? Using more risk balancing in portfolio construction methodology will improve the overall way of how you build portfolios. 

The third one is about risk management, creating a much more dynamic approach. You cannot assume relationships remain the same over a long period. You cannot assume your stakeholders will allow you to stick for the next 10 to 15 years with the portfolio you have. You run the risk, if your drawdown or intra-horizon volatility is too high, that your stakeholders will tell you to change investment policy, and you lose control over the portfolio. 

This all has to be balanced against the objectives of the client. You need to agree with the client: What are their return expectations? Are they realistic? And ask the client: Do you understand what the consequences of risk are? 

Over the last 30 years, we have been in a return-driven environment – we’ve only talked about return and return opportunities. We haven’t made our clients well aware of the risk that a certain investment policy is leading to. In the crisis we weren’t killed by active managers. In the crisis we were killed by beta. 

Active managers did not underperform by 60 per cent. It was equity markets in some areas that went down by that degree. And that is conceptually what people need to understand.

Yuri Bender: And moving into the final segment of our discussion, the shape of tomorrow’s wealth and asset managers: what are organisations going to look like? 

Amin Rajan: Any organisation, no matter what its business model is, if it puts client interests first, it stands a much better chance of success. 

Our work also suggests we are likely to see a lot of fragmentation in the industry. Digitisation is going to have a major impact, particularly at the commoditised end. We’re seeing a vast cohort of post-war baby boomers retiring. Their balances are not significant enough for them to rely on expensive advice, so they’re going to rely on DIY options, and we’re going to see a marriage made in heaven between technology and ETFs. It’s already happening in the US and now it’s happening here in Europe. 

And more and more asset managers are beginning to realise that, as well as thinking about having skilled managers and so on, they also really need to think about implementation issues. 

I think it was Bill, who mentioned 80 per cent of returns came from asset allocation. A study we published last year showed that figure has dropped to 50 per cent. And the other 50 per cent comes through implementation expertise. And implementation in the institutional space basically means: how good is the governance structure? So any asset management houses with really good implementation, they’re really likely to do well. 

My final point is really about asset managers who are able to manage client expectations about relative returns and absolute returns. What we’ve seen is huge client interest in uncorrelated absolute returns. So as a result, absolute returns strategies have grown. But as you know and I know, absolute return strategies tend to underperform in an up market. And in the up market, clients want relative returns, despite the fact they’ve signed up for absolute return mandates. 

This is where client education really needs to be improved, because we are seeing unnecessary tension between managers and clients, in that clients basically want absolute returns in a down market and they want relative returns in and up market, and they can’t really have their cake, eat it and see it grow at the same time. 

Daniel Kuehne: It is a big opportunity for Swiss banks, with their tradition of being close to clients and trying to understand their needs.

That’s not an easy thing in itself, as we need to understand targets and risk – in terms of both how much risk they can tolerate before exiting a strategy and also willingness to take risk. 

This is the first and most important function that private banks will have.

The second is concentrating on asset allocation, something wealth managers need to have a hold on to understand how to structure portfolios, whether traditional or non-traditional, active or passive. These two levels form the core task of the wealth managers. But there is also a third – the industrialisation of producing these building blocks within the relevant concepts.

Pascal  Blanqué: To a large extent, the way you are organised reflects the way you are thinking, and the opposite is true. You think, often, the way in which you are organised. 

We are all talking about risk factors, etc, but we are still organised across asset classes, passive vs. active, those kinds of things. Secondly, we’ve got to understand. To get a better understanding means we’ve got to invest more in understanding, and to accept failure as part of the game. There is no one-size-fits-all solution across the investment space – success in our space is a function of failure. This is a managerial and an investment issue.

Bill O’Neill: I still think this idea of a systematic approach is extremely important. It’s not synonymous with unquestioned acceptance of modern portfolio theory in a big block; it still houses many approaches. Diversification is still extremely important as a tenet, the Holy Grail. 

Yuri Bender: We’ve looked at how the demise of modern portfolio theory has particularly been affected by the environment, how this will change the structure of asset management groups, how it will lead to organisational change, including distribution, and the changing nature of the distributor. We’ve looked at how these changes will affect asset management techniques, in terms of asset allocation in particular. And most of all, I think the message coming home is that the new structures and ways of looking at assets have to really focus around the client and the needs of the client.