RISK FACTOR PORTFOLIO MANAGEMENT WITHIN THE ADVICE FRAMEWORK Putting client needs first Risk means different things to different people. Everyone is exposed to risks of various types inflation, injury, death, longevity, interest rates, political, economic, capital market, wages, liquidity and they are unique to each individual. They arise from the changing interaction between a person s human capital resources, and the future income needed to support their required and desired levels of consumption. At its core, this is the concept of asset-liability matching that everyone is familiar with. If you have a known liability in the future, it is the act of setting aside an asset with a future cash flow matched to immunize against this liability. The role of a portfolio is to immunize the client against their future consumption requirements and the risks associated with them. Simply investing in a mix of defensive and growth assets based on how much the client can handle the market falling in any one year is very far removed from this concept it has little to nothing to do with what they are actually trying to achieve. We need to bring portfolio construction back to the client what are their future liabilities, and what risks should we (and what shouldn t we) expose them to? Not only do we need to understand how much risk a client may want to take on, we also need to assess their capacity to take on risk. Unfortunately in most circumstances clients will need to make sacrifices and bear certain risks that they perhaps wish they didn t have to. It is rare for a client to have enough financial capital to fund all their future liabilities no matter the circumstances, so often trade-offs need to be made. To address this, we propose a framework that no longer focusses on asset classes and future expected returns, but instead on risk factors and the distribution of future returns (i.e. the uncertainty around future expected returns). Risk factors in the Financial Advice process We can broadly categorize the risk factors that clients are exposed to in their daily lives into two buckets: 1) those relating to human capital (current and future income); and 2) liabilities (current and future consumption). Some examples of each are: 1) Human Capital: Wage inflation, Employment, Mortality, Disability 2) Liabilities: Future consumption, Goods and services inflation, Longevity, Liquidity, Taxation These risks play a special role in personal financial management, as their exposures are created independently of any particular investment strategy. That is, they are created as a result of living, regardless of whether an individual has any financial capital that requires management. As individuals, we are very familiar with the above risk factors, our exposure to them, and ways in which we can mitigate them. We typically manage human capital related risks directly ourselves, by investing time, money and effort into education and maintaining physical health. This is particularly the case for mild stresses. However when severe stresses are considered, most people have little risk appetite (both capacity and tolerance) to bear them. Thus they choose to transfer these risks, typically via insurance such as life, disability and health insurance. Consumption related risks may be similarly managed through insurance strategies such as home and contents insurance to mitigate physical property risk, or superannuation to manage longevity risk. Banking solutions such as fixed rate loans and mortgages help mitigate interest rate risk. However many consumption related risks cannot be so easily managed or mitigated, such as the various types of inflation risks (e.g. consumer, rental, property prices), liquidity, taxation and political risks. Joshua Corrigan and Dan Miles Page 1
Some of these risks are driven by the economy and capital markets, whilst others are driven by the political system (a derivative of the social, demographic and behavioural systems). In the Advice Process it is critical to identify those liability risk factors that relate to the capital markets, either directly such as interest rates, or indirectly through say the economic impact on property and rental price inflation. This is because it is necessary to understand whether an individual s exposures to such risks are within their risk tolerances or not. If they are, then they can bear more of this risk through investment exposure, but if not, they will be averse to bearing additional amounts of this risk, and may actively seek out ways to reduce it through various investment strategies. A risk factor approach to portfolio construction With that framework in mind, it is easy to see how traditional portfolio construction methodologies fall short. There is little to no relevance given to the risk factors a client is exposed to, and what they can do with their financial capital to mitigate them. Given that clients have human capital, liabilities and the risk factors associated with them both how does financial capital, or assets, fit into this? Quite simply it goes back to asset-liability management. Financial assets are used to reduce and/or manage the risks associated with human capital, and to help fund current and future liabilities as well as the uncertainty surrounding them. Asset classes are largely irrelevant. Asset return behavior is driven by a number of different risk factors. It is these factors that are relevant, not the asset class itself 1. Our proposed framework involves breaking down asset classes into their return drivers, understanding the level of influence of those return drivers and the risks (or potential outcomes) associated with them. As an example, below is a table of asset classes mapped to various risk factors: Fixed income Asset Classes Real Cash Actual Inflation Expected inflation Duration Credit / Default Liquidity Geographic / Political Official cash ü ü Bank Bills ü ü ü Inflation swaps ü ü Inflation indexed govt bonds ü ü ü Nominal govt bonds ü ü Inflation indexed corporate bonds Nominal investment grade corp bonds ü ü ü ü ü ü ü ü High yield bonds ü ü ü ü ü Emerging sovereign bonds Global investment grade bonds hedged ü ü ü ü ü ü ü ü ü ü ü Global high yield ü ü ü ü ü ü 1 Please refer to A New Approach to Portfolio Risk Management - The Use of Risk Factors to Rebuild Portfolio Construction Joshua Corrigan and Dan Miles Page 2
Equity Asset Classes Dividend Yields Earning Growth Valuation (P/E) Sector / Industry Style Factors Geographic / Political Currency Dividend swaps ü Index Futures ü ü Market index ü ü ü Large / small cap, Growth / value, momentum ü ü ü ü Sector index ü ü ü ü Developed markets ü ü ü ü ü Emerging markets ü ü ü ü ü Other Asset Classes Cashflow Yields Inflation Earning Growth Valuation (P/E) Liquidity Geographic / Political Property ü ü ü ü ü Infrastructure ü ü ü ü ü ü The output of this process provides us with an expected return payoff profile for each asset class, but more importantly, a framework to understand the inter-relationships across those asset classes. This is important, because it allows us to test any proposed portfolio against those risk factors that clients have particular susceptibility to. As part of the advice process, understanding a client s tolerance to the risks they are exposed to is fundamental. So the role of portfolio construction is to now incorporate the client s risk tolerance and expose them to those future payoff profiles that can help mitigate the risks that they are exposed to via their human capital and liabilities, within the level of tolerance they are comfortable with. We are now no longer talking about exposing clients to asset classes, but to particular potential risks and payoffs again, asset classes are actually largely irrelevant 2. To do this, obviously we must start with an idea of future returns and the distribution, or risks, associated with them. Second, we must understand the inter-relationship of risk factors across the investible universe from which we can build a portfolio solution. 2 We recognise that clients will have particular biases or desires to have exposure to certain asset classes. This introduces a level of complexity, but this is a minor issue. It may greatly skew the results a client will receive, but at the end of the day, it is still the choice of the client and so must be worked with Joshua Corrigan and Dan Miles Page 3
Relating this to client examples To illustrate how this works, we ll take a single client circumstance, but change one element their financial wealth. The client details are below: A couple aged 65 who no longer work Own their own home, debt free Have the following objectives, in order of preference: 1. Minimum income requirement of $3,000 per month (indexed to CPI) 2. An Aspirational income of $4,500 per month (indexed to CPI) 3. Access to $100,000 of liquidity in case of emergency 4. They would like to be able to leave $100,000 (in addition to their family home) to their children from the age of 80 onwards We ll give these clients 2 different levels of financial wealth: Client A = $500,000 in assets (excluding the family home) Client B = $1,250,000 in assets (excluding the family home) Given their future consumption requirements, an indicative value estimate of these liabilities is: Minimum income = $920,000 (Present Value of an inflation indexed annuity) Additional aspirational income = $460,000 (Present Value of an inflation indexed annuity paying the marginal amount of $1,500 per month) Liquidity buffer = $100,000 (current value) Financial bequest = $60,000 (Present Value of $100,000 in 15 years) These two clients have a series of liability risk factors: real interest rates, inflation, longevity and liquidity. The question then becomes how much financial capital (assets) they have to deal with these risk factors, and hence their risk capacity. The current value of the client s future liabilities is shown in the figure below compared to their current total financial capital: Joshua Corrigan and Dan Miles Page 4
From the above it can be seen that client A can t afford their minimum income requirements based on their current assets. In other words, to just maintain their minimum lifestyle requirements they can t simply buy an annuity, term deposit or leave all their money in cash and simply live off the interest. This is the reality of the situation for most clients. Client B on the other hand has enough to fund their minimum income requirement (although they still arguably have some residual real interest rate and inflation risk), but not enough to achieve any of their other goals. Client A has the option of choosing underfunded certainty they know they can t fund their future requirements if they adopt a low capital market risk appetite level (for risks such as economic, credit, valuation, momentum, liquidity). Instead they may choose to accept certainty around some liability risks such as inflation, and uncertainty by bearing others such as longevity and credit risk. It isn t a great outcome, but at least it s a relatively certain one. Their other choice is potentially a better outcome but with uncertainty they can take on capital market risk factors that provide a higher expected return payoff profile, but with greater uncertainty. Client B is in the envied position of being able to fund their minimum lifestyle requirements from their financial assets however, they can t fund any of their other goals, but they do have $270,000 of free capital remaining. In each case the decision is not clear, nor is there only one solution for each client. Client A must make some sacrifices give up some certainty to potentially achieve more of their objectives, or take on more certainty but only partially achieve their goals. Client B can immunize a minimum income requirement and use their remaining $270,000 to try and fund other objectives but reality is rarely this cut-and-dry. Whilst a $3,000 weekly income might be their highest priority, having some money to fund some aspirational liabilities, or leave to their children, might start muddying the waters of the client s objectives. They may be prepared to live with some uncertainty around the $3,000 weekly income to increase the likelihood that they could achieve their aspirational goals, liquidity and bequest objectives. Clearly, this is the role of the Financial Adviser to flesh this out for each client and tailor the outcome specifically for them. More on this later. What a risk factor portfolio would look like for different client examples We must now look at what can be done with the client s financial assets to best try and immunize, or match, against their future liabilities. However, it is worth noting again that liability risk factors are unique in that their exposures are created independently of any particular investment strategy. We have previously stated that asset classes are irrelevant to this process, it is the risk factors and their potential payoff distributions that matter most, so we will look at three overly simplified potential portfolios. However, these portfolios are still comprised of investment building blocks (asset classes, securities, derivatives etc.). As an example, let s look at five potential investment strategies as building blocks. These will change dynamically as conditions evolve, and we are using a small sample size for simplicity. The following diagram shows the expected (ex-ante) annual return distribution for each of these risk factor investment strategies. Joshua Corrigan and Dan Miles Page 5
Ex-Ante Risk Factor Universe These five building blocks can be combined in various ways to manipulate a forecast payoff profile or distribution tailored to the types of tolerances and stresses that are relevant for the client. To illustrate further we propose 3 desired future payoff profiles: 1. High certainty, annuity-type payoff. The skinny portfolio Joshua Corrigan and Dan Miles Page 6
Performance and uncertainty Mean Standard Deviation Prob. Negative 10 th Percentile 3.0% 1.0% 0.3% -0.7% Risk Factor Exposure Cash Gov t Debt Equity Option Vol. Protection ü ü - - - 2. Higher expected return, medium uncertainty with truncated downside. The skewed portfolio Joshua Corrigan and Dan Miles Page 7
Performance and uncertainty Mean Standard Deviation Prob. Negative 10 th Percentile 5.6% 4.5% 6.9% -7.1% Risk Factor Exposure Cash Gov t Debt Equity Option Vol. Protection ü - ü ü ü 3. High Uncertainty, High potential return. The fat portfolio Performance and uncertainty Mean Standard Deviation Prob. Negative 10 th Percentile 6.0% 8.4% 23.5% -20.8% Joshua Corrigan and Dan Miles Page 8
Risk Factor Exposure Cash Gov t Debt Equity Option Vol. Protection - ü ü - ü All of the above portfolios could comprise any mixture of risk exposures or number of risk exposures. For illustrative purposes we have elected to limit those numbers here. However, in a dynamic world, the risk and potential payoff of all those risk factors will change throughout time. Therefore, the makeup of these portfolios will change as well. Viewed together, their return distributions look like the below: Fat, Skinny and Skewed Along the bottom (x) axis is return, across the side axis (y) is the marginal probability of achieving each return segment. These three portfolios are illustrative only. Questions immediately arise when we attempt to make these relevant for client specific circumstances. How much downside can a client reasonably handle? 5%, 10%, 15%? This would determine the nature of the skewed portfolio. How high an average return can we get out of the fat portfolio? This would determine its level of aggressiveness and the size of the fat tails i.e. how big can the worse-case scenarios actually get, what risk factors are driving these worst case scenarios, and how high an average return could we achieve to be rewarded for bearing risk? Joshua Corrigan and Dan Miles Page 9
It is these questions that begin to tailor the output for the specific client, and determine the risk factors that are relevant and should be used to achieve these objectives. It may also include the exclusion of various risk factors clients may not be able to handle any form of inflation risk, and so may need a portfolio immunized against inflation for instance. Translating this for client A and B Given these three simple portfolios, how would they translate to client A & B? This is where the advice process comes into its own. Client A is most likely going to need some form of skewed portfolio. But what is their capacity to take on risk? How skewed does it need to be? Are there any particular stresses that will particularly affect them in the future given their future liability risk factors? For client B there are two likely scenarios. They could use the skinny portfolio for $930,000 of their assets to ensure that their minimum income requirements are met. They may then wish to invest the remaining $270,000 into a fat portfolio to try and meet their other goals. We then need to test to ensure that they re immunized against future liability risk factors (such as longevity and inflation). Alternatively, client B may want some form of skewed portfolio, where their minimum income requirements may be flexed a little to allow for some more potential upside i.e. they are prepared to take on more uncertainty in income for a greater expected return. Again, how much uncertainty? And does this introduce exposure to risk factors within their financial assets that they shouldn t (or don t want to) have exposure to? Stress Testing Stress testing is the manner in which we can test the robustness of proposed portfolio solutions to various risk factor outcomes to gauge a client s level of exposure, their willingness and capacity to take on risk, as well as the suitability of the portfolio to meet these requirements. Below we have stressed two common risk factors economic growth and inflation: Scenario 1 Recession As can be seen from these tests, the fat portfolio, or portfolio C is highly susceptible to economic growth. In the case of a recession this portfolio will perform very poorly. Does the client have the capacity to take on Joshua Corrigan and Dan Miles Page 10
this risk? 3 Are there other areas of their life (liabilities or human capital) that are exposed to economic risk? In the case of clients A and B, that might be the value of their home although, in this case, they simply want to leave it to their children, its value is largely irrelevant, so arguably not. In the case of client A, this skewed portfolio is not well protected against an economic downturn in the short term. Is that tolerable for them in light of the fact that they do not have enough free capital to fund their minimum income requirements? Is an economic downturn something that could hit their lifestyle in a negative way elsewhere (such as increasing the likelihood that they may need to help their children deal with a poor economic environment)? It may be that an economic downturn has little effect on them other than their financial capital, and so is a risk factor that is well worth taking on because it immunizes against other factors (inflation [see scenario 2]; longevity; real rates; etc). Scenario 2 High Inflation In this test, we see that the skewed and fat portfolios fair well, with the skinny portfolio fairing moderately well. For client A, the skinny portfolio is unlikely to be an option, though inflation risk might be a risk factor they have a lot of exposure to. Owning a portfolio such as the skewed portfolio (portfolio B) that is resilient to inflation could be very valuable as they may have a series of future liabilities susceptible to inflation risk. For client B, if they choose to fund their minimum income requirement from a skinny portfolio, the question is then whether their future liabilities could be exposed to inflation risk and if they are, is this skinny portfolio sufficiently inflation resistant to alleviate that risk factor? Conclusion We have been deliberately general in this article as to what answer is best for each client, as well as the make-up of each of these portfolios. The reason for this is that the answer is not black and white it requires the input, skill and professionalism of the adviser to determine the appropriate portfolio. Also, the makeup of these portfolios will change, and can be made up of a series of risk-factor combinations to suit the appropriate circumstances of clients. However, the answer may be as simple as Portfolio B above. Or it could be a combination of portfolios for both client A and B. This is not very different to the strategies that some advisers take with their clients under the current Advice Framework. The differences, we believe, are the following: This process starts with the client first and is focused on an outcome designed for them based on their future needs at an intimate level 3 Please also keep in mind that in a risk-factor framework, there is the ability to build fat type portfolios with lower exposure to economic risk factors. Please refer to A New Approach to Portfolio Risk Management - The Use of Risk Factors to Rebuild Portfolio Construction Joshua Corrigan and Dan Miles Page 11
The focus is on risk management and liability matching. Not only the clients risk tolerance, but also their risk capacity It is grounded in robust testing and modeling, not simply based on rules-of-thumb It aids in building the professional framework within which the adviser can operate, and helps them define their role and demonstrate the value they add in the process. No investment strategy is right for any single person it needs to be tailored for the individual this process brings that back to the heart of the advice process. Appendix Whilst we have provided the illustrative portfolios, one might rightly ask what they would actually be invested in? Otherwise, they re simply theoretical. Below we have the constituents of the three portfolios at an asset class level, but would like readers to note that the constituents, whilst based on forecasts developed as of July 2013, will be very different at different points in time. Makeup of the 3 portfolios: skinny 60% cash +/- 5%; 40% Government Bonds +/- 5% skewed 5% cash +/- 5%; 25% Equities +/- 5%; 70% Protection +/- 10%; 25% notional exposure to short option volatility strategy (a market neutral position, essentially selling market options to generate income) fat 10% Government Bonds +/- 5%; 20% Protection +/-5%; 70% Equities +/- 10% Disclaimer: The above information is commentary only. It is not intended to be, nor should it be construed as, investment advice. The views expressed are subject to change at any time based on market and other conditions. To the extent permitted by law, no liability is accepted for any loss or damage as a result of any reliance on this information. Before making any investment decision you need to consider your particular investment needs, objectives and financial circumstances. Joshua Corrigan and Dan Miles Page 12