The challenge of client-focused investment is coming into starker focus as packaged investment solutions such as model portfolios and target date funds are becoming increasingly popular
The fact an investor’s cashflow requirement or liability forms the basis for any investment strategy is often overlooked or misunderstood.
For a defined benefit pension fund, this cashflow is usually a stream of salary-linked pension payments to retired ex-employees. An insurance company operating with-profits business may need to make a series of guaranteed payments to policyholders at specified dates. A 60-year-old retail investor may use their accumulated fund to generate an annual income of 5% of initial capital over a retirement period of 20 years. A 30-year old might regard their investments as a rainy-day fund.
There are many other examples, but the important common factor is these retail and institutional investments all incorporate some defined cashflow or liability.
Volatility: uses and abuses
Given the cashflow requirement or liability is integral to any client-focused investment strategy, it seems strange most investment processes continue to rely on risk measures such as volatility.
Volatility and tracking error are used to assess the risk a fund manager has taken in order to generate return or ‘alpha’ within a portfolio. Since cashflows are usually outside the fund manager’s control, these time-weighted risk measures have been deliberately designed to remove the impact of the size and timing of cashflows. However, the use of volatility and tracking error has been extended into many aspects of the investment process, including asset allocation, product design and fund rating.
While volatility or tracking error may be useful tools for assessing the past performance of an active fund manager, ignoring the cashflow or investment objective is liable to be inappropriate for the purpose of designing client-focused investment solutions.
Client cashflows create path-dependent risks
Let’s go back to our 60-year old retirement investor: we will assume he has accumulated investments of £100,000 and wants to draw down an income of £5,000 at the start of each year. Assuming the investments generate a fixed annual return of 5% (the lower rate of return in a standard FSA deterministic pension projection), these will meet his income requirements – there would be around £90,000 remaining after 20 years of withdrawals. But no investment fund returns exactly 5% every year. In Figure 1, we compare the fixed 5% return with two alternative 20-year scenarios.
Scenario 1: Fixed 5% annual return (“lower rate” FSA deterministic illustration).
Scenario 2: 10% return for first 17 years, 20% fall for last three years (bad return sequence in last three years).
Scenario 3: 20% fall for first three years, 10% return for last 17 years (bad return sequence in first three years).
The key point is the ‘time-weighted’ return (total return over the 20-year period) is the same in all three scenarios – exactly 5%. The difference between the three scenarios is the sequence of returns.
In the absence of any cashflows, these three scenarios would produce exactly the same outcome: the initial fund of £100,000 would grow to £260,000 after 20 years. However, the fact the investor is drawing an annual income of £5,000 means the sequence of returns has a dramatic impact on the outcome. While the income can be comfortably supported for 20 years under a fixed deterministic return, under scenario three the fund runs out completely after 17 years.
There are a few important points to take away from this example:
- When selecting an investment strategy or product to fit the client’s risk profile, we should use a measure of risk which accounts for the fact the outcome depends on the sequence of returns.
- The total return over the period – eg whether it is 5%, 7% or 9% – is much less important.
- Customers and advisers relying on time-weighted risk measures or deterministic return scenarios are likely to significantly underestimate the risk in a given investment strategy.
- Conventional time-weighted risk measures like volatility or tracking error are not useful for designing or selecting an appropriate investment strategy given this type of cashflow profile.
A fund-rating paradox
Having used a scenario-based approach to select an investment strategy that matches the client’s investment objective and risk profile, fund ratings may be used to help select funds.
Generally, these fund ratings will use historic returns to estimate various metrics such as expected return, volatility, risk-adjusted return, and for active funds a tracking error versus a defined benchmark.
Let’s assume our retirement investor had decided to invest 100% of his assets in UK equities. Suppose he then had a choice between the three funds listed in Figure 2.
On the face of it, this looks like a ‘no-brainer’: Fund 2 is outperforming the index and has a lower tracking error. Fund three is underperforming the index and has a higher tracking error.
Faced with this information, the ratings are intuitive and most investors would immediately rule out Fund 3.
But going back to our previous analysis, we saw the client cashflow was most susceptible to a bad sequence of market returns over the first three years of the 20-year investment term – Scenario 3. In this scenario, Fund 2 will underperform over the first few years, while the defensive fund actually delivers slightly higher returns.
The outcomes for the investor are shown in Figure 3. Given the investment objective, Fund 3 is likely to be a better choice than either Fund 1 or Fund 2. Without taking client cashflows into account, the fund data would have led to a very different answer!
Application of scenario-based asset models
In practice, scenario-based asset liability modelling allows us to calculate risk measures that take into account different possible sequences of returns when designing client-focused investment solutions. There are two methods for creating economic scenarios (see Figure 4).
Implications for investment management
This challenge of client-focused investment is coming into starker focus as packaged investment solutions such as model portfolios and target date funds are becoming increasingly popular. These propositions are targeted at particular customer needs and as a consequence embed some element of advice. The marketing message is moving away from pure performance to robust governance.
From a regulatory perspective, the FSA has repeatedly emphasised the need to ensure products are designed in accordance with customer needs through its increasingly robust enforcement of Treating Customers Fairly.
Scenario-based models offer a reliable tool for designing investment products that do what they say on the tin. Over the last 12 months, we have seen the market waking up to this challenge – many of our partners and clients have committed significant resource to embedding scenario-based modelling into their product design and communication processes.
However, those organisations represent a small number of market leaders – there is still a long way to go before the majority of UK investors will experience truly client-focused investment products.
Phil Mowbray, head of product risk at Barrie & Hibbert
Figure 1:
Figure 3:
