Managing investment outcomes with volatility control Andy Connell Co-Head of Portfolio Solutions Swings in asset prices are a concern for most investors. Some, however, suffer more than others. Many want the returns that equities can give, but cannot live with the risk created by short-term market swings. One answer is to use a volatility-controlled strategy. Here we outline how a range of such techniques can reduce the risks, while retaining much of the gains to be had from risky assets like equities. We argue that, while the results will inevitably be different, the risk-adjusted returns should be much improved. Mike Hodgson, Ph.D. Head of Risk Managed Investment & Structuring Approaches to portfolio construction vary considerably across investor types. However, when assessing an allocation to an asset class or a multi asset portfolio most investors will give some consideration to the distribution of possible outcomes. Often, investors will use medium term historic volatility as an estimate for future volatility to estimate the distribution of future returns. However, volatility itself can vary considerably. Volatility control techniques aim to target or limit the volatility of portfolio returns over time by adjusting exposure according to the near term volatility forecast, thereby seeking to control the distribution of final outcomes. These techniques have been generally successful because empirically recent historical volatility has been a good indication of near term volatility. Volatility control techniques can be an effective part of a tailored strategy seeking to achieve a number of investment objectives This paper examines the effect of volatility control on the distribution of returns, expected returns and investment outcomes and in so doing specifically aims to address common misperceptions about these techniques. Specifically it examines three techniques the volatility target, the volatility cap and the variable volatility cap (or V VC) in the light of several key questions. ȂȂ ȂȂ ȂȂ What is the underlying investment insight that volatility control techniques aim to exploit? What are the volatility control investment techniques? What is the effect of the investment techniques over the long term on investment outcomes? ȂȂ What is the investor experience over the short and medium term and therefore which types of investors find these outcomes attractive? The underlying investment rationale Today s volatility gives an insight into tomorrow s volatility Underpinning all volatility control techniques is the premise that recently experienced volatility tells us something about tomorrow s volatility. This is not a new concept or insight: there is a wide range of empirical studies that support this premise. Robert F. Engle and Tim Bollerslev s work resulted in the development of the GARCH (Generalized Autoregressive Conditional Heteroskedasticity) model. This showed that volatility clustering occurs in equity markets (i.e. large changes tend to be followed by large changes, of either sign, and small changes tend to be followed by small changes). While it is always reassuring for an investment technique to be supported by a Nobel prize 1 it is just as important to re-confirm the basis for our investment approach. Figure 1 shows the range of the next day s returns in excess of the short term interest rate (on the vertical axis) against the volatility derived from recent returns on the horizontal axis, (with lower volatility on the left and higher volatility on the right). Looking at the horizontal scale, we immediately see that volatility can take a wide range of values. The solid vertical blocks show the range of the next day s returns that are one standard deviation either side of the mean, i.e. one day s volatility. The vertical lines show the full extent of observed returns, lowest to highest. 1 Robert Engle and Clive Grainger were awarded the 2003 Nobel prize in Economic Sciences for methods analysing economic time series with time varying volatility (ARCH) 1 Managing investment outcomes with volatility control For Financial Intermediary, Institutional and Consultant use only. Not for redistribution under any circumstances.
The Figure shows that as the recent measured volatility increases (from left to right) so does the volatility of tomorrow s returns, (as measured by the height of the solid bars). Figure 1: Comparing the breadth of tomorrow s returns with recent measured volatility Distribution of next-day returns 2 15% Future volatility increases as historic volatility increases 1 5% -5% -1-15% -2-25% 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 Measured volatility % Source: Schroders, Bloomberg, Federal Bank of St. Louis. Daily data December 31, 1927 to December 31, 2012. Indices used are the S&P 500 Index (adjusted for dividends by Schroders) (January 1963 to January1988), and the S&P 500 Total Return Index (January 1988 to December 2013). These are the results of a backtest. Certain assumptions have been made. Actual results would vary. In Figure 2, we plot the magnitude of the observed standard deviation of daily returns against the recent volatility measure and compare it to the forecast oneday standard deviation of returns derived from that volatility measure. We can see how well the volatility measure predicted the standard deviation of the future daily returns over a wide range of volatilities. This is evidence that as far as volatility is concerned, recent experience tells us about the near future. Figure 2: Comparing forecast and actual volatility 2.5% 2. 1.5% 1. 0.5% 0. 5% 1 15% 2 25% 3 35% Realized Volatility Forecast Volatility Source: Schroders, Bloomberg, Federal Bank of St. Louis. Daily data December 31, 1927 to December 31, 2012. Indices used re the S&P 500 Index (adjusted for dividends by Schroders) (January 1963 to January1988), and the S&P 500 Total Return Index (January 1988 to December 2013). The population of the daily measures of recent volatility have been split into 5% bands and plotted the standard deviation of the next day return against the average volatility within each 5% band. These are the results of a backtest. Certain assumptions have been made. Actual results would vary. Managing investment outcomes with volatility control 2
Figure 3 below illustrates how we might use this observation to aim to control the volatility of returns of a portfolio. It shows the results of implementing a volatility target. While this technique is explained in more detail later on, at this point it is sufficient to understand that a volatility target aims to deliver a portfolio with a constant volatility by reducing exposure during higher volatility periods. The graph shows the breadth of return outcomes from a portfolio implementing a volatility target strategy with a one day lag 2 on the vertical axis against the daily estimates of volatility derived from recent returns on the horizontal axis. The graph shows that the volatility target strategy delivers an almost uniform breadth of returns, (as measured by the height of the solid bars), across the range of recent measured volatility. This gives some evidence that by using recent volatility we can aim to control the standard deviation (i.e. the volatility) of returns of a portfolio in the future, and hence the distribution of returns over an investment horizon. Figure 3: The volatility of returns when implementing a volatility target strategy Distribution of allocation-weighted next-day returns 1 5% -5% -1-15% 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 Measured volatility % Source: Schroders, Bloomberg, Federal Bank of St. Louis. Daily data December 31, 1927 to December 31, 2012. Indices used are the S&P 500 Index (adjusted for dividends by Schroders) (January 1963 to January1988), and the S&P 500 Total Return Index (January 1988 to December 2013). These are the results of a backtest. Certain assumptions have been made. Actual results would vary. Is higher volatility rewarded with sufficiently higher returns? The charts above suggest that recent volatility tells us about tomorrow s volatility and we can apply this insight to seek to control the volatility of a portfolio s returns by reducing exposure during higher volatility periods. It is natural to consider how this strategy may affect returns. To explore this we have looked at the average experienced excess return versus forecast volatility, (Figure 4). While we saw in Figure 2, a very clear empirical relationship between the next day spread of returns and measured volatility, there is no clear pattern between the next day average excess return and measured volatility. In particular, it is not clear that higher volatility is being rewarded with higher expected return. We should therefore expect that volatility control, which reduces exposure during higher volatility periods, will improve the expected returns per unit of risk. In fact, the historic data suggests that as volatility rises, average future returns do not increase enough to compensate for the extra risk. There is a common misconception that proponents of volatility control advocate that volatility can be used to forecast returns, and/or that heightened volatility results in negative returns. Rather, this and other studies 3 show that the relationship between historic volatility and future expected returns is not clear. In fact, the historic data suggests that as volatility rises, average future returns do not increase enough to compensate for the extra risk. 2 The reason for implementing a lag in the back test is to offer some evidence that the ability to manage the breadth of returns (i.e. volatility) of an investment strategy is robust to a reasonable implementation delay. 3 Volatility Managed Portfolios, Moreira and Muir 2016. 3 Managing investment outcomes with volatility control
Figure 4: Future returns vs. experienced volatility 0.07% 0.05% 0.03% 0.01% -0.01% -0.03% -0.05% 5% 1 15% 2 25% 3 35% Average Return Source: Schroders, Bloomberg, Federal Bank of St. Louis. Daily data December 31, 1927 December 31, 2012. Indices used are the S&P 500 Index (adjusted for dividends by Schroders) (January 1963 January1988), S&P 500 Total Return Index (January 1988 December 2013). The population of the daily measures of forecast volatility have been split into 5% bands and plotted the average level of the next day return against the average volatility within each 5%. The gray line represents a level. These are the results of a backtest. Certain assumptions have been made. Actual results would vary. Rather, this and other studies show that the relationship between historic volatility and future expected returns is not clear. Volatility control for long-term investors It is often stated that long-term investors should be less concerned by short-term volatility given they have time to ride out short term setbacks (even when these are extreme). This view is often also combined with the belief that long-term investors shouldn t be reducing their exposure after periods of extreme loss (perhaps they should even be increasing their exposures) because it is at this point that markets offer most value. To understand whether a volatility control philosophy can inform this approach, we examine the volatility, return and risk return ratio in the period immediately after a market shock. Over the period there are 37 drawdowns of 15% or more. Figure 5 shows the average volatility and average return in each of the 10 months following the drawdown. These are expressed as a percentile relative to the 50th percentile of their respective histories. So an observation of +25% is a high reading (equating to the 75th percentile) and an observation of -25% is a low reading (equating to the 25th percentile). It highlights that, on average, volatility is highly elevated throughout the period and that after the drawdown, returns are disappointing relative to the median. This analysis appears to dispel the assertion that there is great opportunity available to long term investors in the period immediately after a large loss. Figure 5: Risk, return and the ratio of return per unit of risk in the period after a 15% fall in the market 5 4 3 2 1-1 -2-3 -4-5 1 2 3 4 5 6 7 8 9 10 Months (after 15% loss) Volatility Return Ratio of return to risk Source: Schroders, Bloomberg, Federal Bank of St. Louis. Daily data December 31, 1927 to December 31, 2012. Indices used are the S&P 500 Index (adjusted for dividends by Schroders) (January 1963 to January1988), and the S&P 500 Total Return Index (January 1988 to December 2013). These are the results of a backtest. Certain assumptions have been made. Actual results would vary. Managing investment outcomes with volatility control 4
The investment techniques So far we have provided the basis for the importance of controlling volatility. By seeking to reduce risk during periods of higher volatility (say, by reducing exposure to equities), an investor may be able to improve the expected returns per unit of risk. In other word, a portfolio would become more efficient. For purposes of this report, we now examine three common outcomes that investors seek to achieve when managing volatility. The Volatility Target This technique aims to manage the volatility of the portfolio so that it remains at a constant level over time. For example, if an equity portfolio is managed to a volatility target of 1 then the process is as set out below. Tomorrow s exposure to the risky asset = 1 (i.e. the target level) / Today s volatility estimate. For example: 1. If the target is 1 and today s volatility estimate is 15% then the allocation to the risky asset is 1 / 15% i.e. 67%. 2. If the target is 1 and today s volatility estimate is 8% then the allocation to the risky asset is 1/8% i.e. 125%. The Volatility Cap The volatility cap aims to modify an investor s experience by measuring the daily volatility of an investment and reducing the exposure to the risky asset if the volatility exceeds a pre-defined (i.e. cap) level. The process used to achieve this is relatively straightforward and is set out below: Tomorrow s exposure to the risky asset = The lower of: 1. A 10 allocation to the risky asset; or 2. The The Volatility Cap level / Today s portfolio volatility estimate. For example: if we have an equity portfolio with a volatility cap of 15% and: 1. Today s volatility estimate is 8%, then as this is below the cap, tomorrow s portfolio will be a 10 allocation to equities. 2. Today s portfolio volatility is 2 and as this is above the cap level, tomorrow s volatility will be (the cap level) 15% / (today s volatility estimate) 2, i.e. 75%. The Variable Volatility Cap (VVC) A variable volatility cap aims to limit the risk of loss below a specific level. Figure 6 gives an overview of how a V VC works. Essentially, the implementation formula is the same for the volatility cap (as set out above) except the level of the volatility cap is variable, dependent on the level of loss the strategy has incurred from the high watermark. As losses increase, the level of the volatility cap falls. This aims to reduce the risk taken in falling markets and so limit the losses incurred by the strategy. Figure 6: The variable volatility cap limit declines as the experienced drawdown grows bad outcomes Maximum loss objective Valatility cap level Capital loss from highest point Source: Schroders, for illustration only By way of example: 1. If the capital loss since the highest point is 7% and the cap level that corresponds to this is 1 then if the risky asset volatility is 12% then the allocation to it within the portfolio will be 1 / 12% = 83%. 2. As the amount of loss increases, the cap level also falls to a minimum level. So for example if the loss level is 12% (i.e. the strategy is only worth 88% of its maximum value) the cap level is then reduced to 5%. If the underlying asset s volatility remains at 12% then the allocation to the risky asset is 5% / 12% = 42%. 5 Managing investment outcomes with volatility control
The effect of volatility control techniques on investment outcomes In this section we will look at the three volatility control techniques and explore their potential effect on the investment outcomes in terms of: Expected excess returns Volatility Drawdown (i.e. capital loss from peak to trough) The ratio of expected excess return per unit of risk The volatility control techniques used for our comparative analysis is set out in the table on the next page. None of these techniques offers any investment guarantee that describes a return outcome relative to the underlying risk asset (i.e. these are not benchmark relative strategies). Although we will compare each strategy to an underlying index, the aim of the comparison is not to show a better return but to show how the volatility control technique may adjust the distribution of investment outcomes. In all three cases, it is important to note that, although a volatility controlled investment is driven on a day-to-day basis by the performance of the underlying asset whatever that may be the ultimate returns will differ due to the allocation to that asset varying over time. Figure 7: The key elements of the volatility control examples Volatility target Volatility cap VVC Level 15% 15% Varies depending on the loss level from the highest point over the last 12 months Maximum exposure 15 10 10 Objective To deliver a (broadly) stable volatility over time To deliver a volatility no greater than the cap level Mitigate maximum level of loss from the highest point to a maximum level of 15% on a rolling 12 month basis Source: Schroders. For illustration only. No investment strategy or technique can guarantee any level of return or protect a portfolio from loss of principal. Figure 8 and Figure 9 illustrate how the application of these techniques on the S&P500 would have affected the outcomes. Figure 8: The distribution of annual returns from the application of volatility control techniques -5-4 -3-2 -1 1 2 3 4 5 6 7 8 >8 S&P 500 TR Index Volatility Cap Volatility Target Variable Volatility Cap Source: Schroders, Bloomberg, Federal Bank of St. Louis. Daily data December 31, 1927 to November 30, 2015. Indices used are the S&P 500 Index (adjusted for dividends by Schroders) (January 1963 to January1988) and the S&P 500 Total Return Index (January 1988 to November 2015). These are the results of a back test. Certain assumptions have been made including the implementation of the volatility control techniques with a 1 working day lag. Actual results would vary. Managing investment outcomes with volatility control 6
Figure 9: The distribution of investment outcomes from the volatility control techniques S&P500 Volatility target Volatility cap VVC Return (p.a.) 8.03% 8.54% 6.98% 6.57% Volatility 18.76% 14.34% 12.58% 10.18% Drawdown -70.09% -43.32% -38.65% -20.08% Excess return/risk 0.43 0.60 0.55 0.64 Source: Schroders, Bloomberg, Federal Bank of St. Louis. Daily data December 31, 1927 to November 30, 2015. Indices used are the S&P 500 Index (adjusted for dividends by Schroders) (January 1963 to January1988) and the S&P 500 Total Return Index (January 1988 to November 2015). These are the results of a back test. Certain assumptions have been made including the implementation of the volatility control techniques with a 1 working day lag. Actual results would vary. The graph and the table are intuitively consistent with our discussion thus far. They suggest that: Volatility management techniques will alter the breadth of investment outcomes. The chart shows how the occurrence of extreme gains and losses are eliminated as a result of the techniques and the table confirms this both in terms of the volatility number but also in terms of the level of loss (drawdown) over a 12 month period For the above parameters, average returns are affected by the volatility control technique As a result of the reduction in volatility and relatively unchanged return numbers, the ratio of excess return to risk improved as a result of each technique As a result of the reduction in volatility and relatively unchanged return numbers, the ratio of excess return to risk improved as a result of each technique. To demonstrate that these results are not dependent on the Great Depression and the Great Financial Crisis of the 1930s and the 2000s, the same analysis excluding these show that the volatility control techniques still resulted in a higher risk return ratio than that of the underlying index. Figure 10: The outcomes from volatility control techniques (excluding 1930s and 2000s) S&P500 Volatility target Volatility cap VVC Return (p.a.) 9.18% 10.36% 8.26% 7.17% Volatility 15.46% 14.27% 12.33% 10.02% Drawdown -46.56% -43.32% -38.65% -20.08% Excess return/risk 0.59 0.73 0.67 0.72 Source: Schroders, Bloomberg, Federal Bank of St. Louis. Daily data December 31, 1927 to November 30, 2015. Indices used are the S&P 500 Index (adjusted for dividends by Schroders) (January 1963 to January1988) and the S&P 500 Total Return Index (January 1988 to November 2015). These are the results of a back test. Certain assumptions have been made including the implementation of the volatility control techniques with a 1 working day lag. Actual results would vary. Investor experience: Implementing volatility control over the short and medium term The previous sections drew conclusions from statistics generated from 88 years of data. However, market conditions vary especially for shorter term periods. Here we consider the difference in outcomes between one of three volatility controlled strategies and the S&P500 for holding periods of a few years. For the purpose of this section, we compare the investment strategies over 12 months. The results for 36 and 60 month periods are set out in the Appendix. In all three cases, it is important to note that, although a volatility controlled investment is driven on a day-today basis by the performance of the underlying asset whatever that may be the ultimate returns will differ due to the allocation to that asset varying over time. 7 Managing investment outcomes with volatility control
In the case of volatility caps and targets, we think such techniques can be very useful in seeking to accomplish the outcomes stated above; however, they are less likely to be suitable for investors seeking explicit downside risk protection Figure 11: The effect of a 15% volatility target in terms of desired outcome and relative returns over 12 month periods Volatility control return 20 6 15 5 10 4 5 3 2-5 1-10 -10-5 5 10 15 20 S&P 500 return 28 33 38 43 48 53 57 S&P 500 Index 62 67 72 77 82 86 91 96 01 06 Year Vol Target Index Target Level 11 Source: Schroders, Bloomberg, Federal Bank of St. Louis. Daily data December 31, 1927 to November 30, 2015. Indices used are the S&P 500 Index (adjusted for dividends by Schroders) (January 1963 to January1988) and the S&P 500 Total Return Index (January 1988 to November 2015). These are the results of a back test. Certain assumptions have been made including the implementation of the volatility control techniques with a 1 working day lag. Actual results would vary. On the left hand side, the chart shows the returns of the volatility targeted portfolio on the vertical axis against the underlying asset s return on the horizontal axis. While this chart indicates that the differential returns between the two can be material (both positive and negative), the right hand side chart shows that the volatility target strategy was reasonably effective at delivering a portfolio with the required volatility level despite the volatility of the underlying asset fluctuating wildly over the period. Figure 12: The effect of a 15% volatility cap in terms of desired outcome and relative returns over 12 month periods Volatility cap return 20 6 15 5 10 4 5 3 2-5 1-10 -10-5 5 10 15 20 S&P 500 return 28 33 38 43 48 53 57 62 67 72 77 82 86 91 96 01 06 Year S&P 500 Index Vol Capped Index Cap Level 11 Source: Schroders, Bloomberg, Federal Bank of St. Louis. Daily data December 31, 1927 to November 30, 2015. Indices used are the S&P 500 Index (adjusted for dividends by Schroders) (January 1963 to January1988) and the S&P 500 Total Return Index (January 1988 to November 2015). These are the results of a back test. Certain assumptions have been made including the implementation of the volatility control techniques with a 1 working day lag. Actual results would vary. Figure 12 illustrates a volatility capped portfolio on the vertical axis against the underlying asset s return on the horizontal axis. Managing investment outcomes with volatility control 8
Similar to Figure 11, this chart shows that the differential returns between the two can be material (both positive and negative). The right hand side chart also shows how the volatility cap strategy was reasonably effective at keeping the volatility below the cap level despite the volatility of the underlying asset fluctuating wildly over the period. The VVC may be suitable for investors who place a high priority on capital preservation, and aims to be a cost effective way to target an explicit maximum loss risk objective Figure 13: The effect of a 15% target drawdown VVC in terms of desired outcome and relative returns over 12 month periods Variable volatility cap return 20 15 10-1 -2-3 5-4 -5-5 -6-7 -10-100 -50 5 10 15 20 S&P 500 return -8 28 33 38 43 48 53 57 S&P 500 Index 62 67 72 77 82 86 Year Variable Volatilty Cap 91 96 01 06 11 Max 1yr Drawdown Source: Schroders, Bloomberg, Federal Bank of St. Louis. Daily data December 31, 1927 to November 30, 2015. Indices used are the S&P 500 Index (adjusted for dividends by Schroders) (January 1963 to January1988) and the S&P 500 Total Return Index (January 1988 to November 2015). These are the results of a back test. Certain assumptions have been made including the implementation of the volatility control techniques with a 1 working day lag. Actual results would vary. Figure 13 shows the results of the third vol-control strategy the VVC. Similar to the results of the other two strategies, these charts show that the differential returns between the two can be material (both positive and negative). However, they also show that: 2. The V VC was effective at mitigating extreme loss, however it was not able in all instances to ensure that the losses over a rolling 12 month period are contained to 15%. 1. The VVC portfolio incurred much smaller drawdowns over the period during which the S&P suffered extreme losses 9 Managing investment outcomes with volatility control
Conclusion We believe volatility control techniques in general can be an effective part of a tailored strategy seeking to achieve a number of investment objectives: Stabilize equity volatility and thereby better control the spread of outcomes ; Limit the cost of options used to provide a capital guarantee or to mitigate losses; Maximize the potential benefits for non-investment reasons, such as Solvency II; Dampen the volatility of a growth portfolio through alternative methods of investing in bonds. What we conclude from our analysis is that volatility control techniques may be appropriate for investors whose criteria for success is consistent with their intended objective. However, index-relative investors should be cognizant that the price of such an altered outcome profile is a very high tracking error (as one would expect since the implementation of the volatility control technique can be expected to achieve different return profiles relative to an underlying equity portfolio). In the case of volatility caps and targets, we think such techniques can be very useful in seeking to accomplish the outcomes stated above; however, they are less likely to be suitable for investors seeking explicit downside risk protection or if performance is to be measured against a conventional market benchmark like an index, as the returns can differ markedly from an indexed portfolio. The VVC may be suitable for investors who place a high priority on capital preservation, and aims to be a cost effective way to target an explicit maximum loss risk objective e.g. to seek to limit potential loss from a strategy s 12-month trailing high water mark to a maximum of 15%. It is important to note that the protection level is on a best efforts basis it is not a guarantee. Therefore, a variable volatility cap may not be suitable for investors who require hard protection. In cases where an investor is seeking more certain protection, a put option or guarantee would likely be more appropriate. Managing investment outcomes with volatility control 10
Appendix 1: The relationship between: ȂȂ ȂȂ Actual/forecast standard deviation of daily returns and recent volatility Average/forecast-risk-adjusted expected excess returns and recent volatility Period 1927 1960 3. 2.5% 2. 1.5% 1. 0.5% 0. 0.2 0.15% 0.1 0.5% 0. 0.5% 0.1 0.15% 0.2 5% 1 15% 2 25% 3 35% 4 45% 5 Period 1960 1990 1.4% 1.2% 1. 0.8% 0.6% 0.4% 0.2% Realized Volitality Forecast Volatility 5% 1 15% 2 25% 3 35% 4 45% 5 Average Return 0. 5% 1 15% 2 25% Realized Volatility Forecast Volatility 11 Managing investment outcomes with volatility control
Period 1960 1990 continued 0.2 0.15% 0.1 0.05% 0.0-0.05% -0.1-0.15% -0.2 5% 1 15% 2 25% Average Return Period 1990 2012 2. 1.8% 1.6% 1.4% 1.2% 1. 0.8% 0.6% 0.4% 0.2% 0. 5% 1 15% 2 25% 3 35% 0.2 0.15% 0.1 0.05% 0.0 0.05% 0.1-0.15% Realized Volatility Forecast Volatility 0.2 5% 1 15% 2 25% 3 35% Average Return Source: Schroders, Bloomberg, Federal Bank of St. Louis. Daily data 31 December 1927 31 December 2012. Indices used are the S&P 500 Index (adjusted for dividends by Schroders) (January 1963 January1988), S&P 500 Total Return Index (January 1988 December 2013). These are the results of a backtest. Certain assumptions have been made. Actual results would vary. 12 Managing investment outcomes with volatility control
Appendix 2: Impact of strategy parameters on final risk and return In the following, we have varied the parameters used for each volatility strategy in order to assess their impact on the final outcome. Volatility Target: The volatility target initially assumed in our paper was 15%. We have re-run the back-test under the same assumptions but using a volatility target level of 12.5% and 17.5%. The outcome of the back-test is in the table below. S&P500 12.5%- Target 15%- Target 17.5%- Target Return (p.a.) 8.03% 7.08% 8.54% 9.63% Volatility 18.76% 12.22% 14.34% 16.16% Drawdown -70.09% -38.75% -43.32% -45.58% Excess return/risk 0.43 0.58 0.60 0.60 Volatility Cap: The volatility cap initially assumed in our paper was 15%. We have re-run the back-test under the same assumptions but using a volatility cap level of 12.5% and 17.5% instead. The outcome of the back-test is in the table below. S&P500 12.5%- Cap 15%- Cap 17.5%- Cap Return (p.a.) 8.03% 6.44% 6.98% 7.34% Volatility 18.76% 11.29% 12.58% 13.58% Drawdown -70.09% -34.11% -38.65% -41.76% Excess return/risk 0.43 0.57 0.55 0.54 Variable Volatility Cap (VVC): The strategy is targeting a maximum drawdown level over a rolling 12 month period. This target level was set to 15% in the paper. We have re-run the back-test under the same assumptions but using a drawdown target level of 12.5% and 17.5% instead. The outcome of the back-test is in the table below: S&P500 12.5%- Drawdown 15%- Drawdown 17.5%- Drawdown Return (p.a.) 8.03% 6.23% 6.57% 6.83% Volatility 18.76% 9.69% 10.18% 10.63% Drawdown -70.09% -18.89% -20.08% -21.18% Excess return/risk 0.43 0.64 0.64 0.64 Source: Schroders, Bloomberg, Federal Bank of St. Louis. Daily data 31 December 1927 to 30 November 2015. Indices used are the S&P 500 Index (adjusted for dividends by Schroders) (January 1963 to January1988) and the S&P 500 Total Return Index (January 1988 to November 2015). These are the results of a back test. Certain assumptions have been made including the implementation of the volatility control techniques with a 1 working day lag. Actual results would vary. Important information on back-tested and/or simulated performance results: Simulated and backtested results in general must be considered as no more than an approximate representation of the portfolio s performance, not as indicative of how it would have performed in the past. It is the result of statistical modeling, with the benefit of hindsight, based on a number of assumptions and there are a number of material limitations on the retrospective reconstruction of any performance results from performance records. For example, it may not take into account any dealing costs or liquidity issues which would have affected the strategy s performance. These and analysis data should not be relied on to predict possible future performance. Indices used herein are widely used, unmanaged proxies for their respective asset classes. Investors cannot invest directly in any index. No management fees are assumed, which would have impacted results. It is important to note that there is not a unique mapping of the outcome of equities to volatility managed equity strategies. Therefore, because a distribution of returns for a particular strategy is considered better, it does not mean that the outcome for that distribution will necessarily be better. For example, over the series of 10 years of returns, the median outcome for the volatility managed equity strategy may be greater than the underlying equity strategy; however, there are scenarios where the equity strategy provides a higher outcome. As with all investment objectives, there can be no guarantee that any investment strategy or technique will achieve their objectives in the future. where the equity strategy provides a higher outcome. As with all investment objectives, there can be no guarantee that any investment strategy or technique will achieve their objectives in the future. Managing investment outcomes with volatility control 13
Appendix 3: Impact of holding period on final risk and return In the paper, we have shown the distribution of the returns for each strategy over 12 month period compared to the S&P 500 returns. We have as well shown that each strategy has achieved its objectives over the same period. In this appendix we have extended the period from 12 month to 3 years and 5 years respectively. Volatility Target: 3 years period Volatility target return 25 20 15 10 5-5 -10-10 -5 5 10 15 20 25 S&P 500 return 6 5 4 3 2 1 30 35 39 44 48 53 57 62 66 71 75 80 84 89 93 98 02 08 S&P 500 Index Vol Target Index Target Level 12 5 years period Volatility target return 40 35 30 25 20 15 10 5-5 -10-10 -5 5 10 15 20 25 30 35 40 S&P 500 return 6 5 4 3 2 1 35 40 45 50 55 60 65 70 75 80 85 90 95 00 05 10 12 S&P 500 Index Vol Target Index Target Level Managing investment outcomes with volatility control 14
Volatility Cap: 3 years period Volatility target return 20 15 10 5-5 -10-10 -50 5 10 15 20 S&P 500 return 6 5 4 3 2 1 30 34 38 42 45 49 53 57 60 64 68 72 75 79 83 87 90 94 98 02 05 09 S&P 500 Index Vol Capped Index Cap Level 13 5 years period VVC: Volatility cap return 40 35 30 25 20 15 10 5-5 -10-10 -5 5 10 15 20 25 30 35 40 3 years period S&P 500 return 6 5 4 3 2 1 32 37 41 45 49 53 57 62 66 70 74 78 82 87 91 95 99 03 07 S&P 500 Index Vol Capped Index Cap Level 12 Variable volatility cap return 20 15 10 5-5 -10-10 -5 5 10 15 20 S&P 500 return -1-2 -3-4 -5-6 -7-8 28 33 37 41 45 49 53 58 62 66 70 74 78 83 87 91 95 99 03 08 12 S&P 500 Index Variable Volatility Cap Max 1yr Drawdown Managing investment outcomes with volatility control 15
VVC continued: 5 years period Variable volatility cap return 40 35 30 25 20 15 10 5-5 -10-10 -5 5 10 15 20 25 30 35 40 S&P 500 return -1-2 -3-4 -5-6 -7-8 28 33 37 41 45 49 53 58 62 66 70 74 78 83 87 91 95 99 03 08 12 S&P 500 Index Variable Volatility Cap Max 1yr Drawdown Source: Schroders, Bloomberg, Federal Bank of St. Louis. Daily data December 31, 1927 to November 30, 2015. Indices used are the S&P 500 Index (adjusted for dividends by Schroders) (January 1963 to January1988) and the S&P 500 Total Return Index (January 1988 to November 2015). These are the results of a back test. Certain assumptions have been made including the implementation of the volatility control techniques with a 1 working day lag. Further details available on request. Managing investment outcomes with volatility control 16
About the authors Andy Connell Co-Head of Portfolio Solutions Andrew is Co-Head of Portfolio Solutions (with John McLaughlin, Ph.D.) at Schroders, and is based in London. Andy joined Schroders in 2007. Previously to working at Schroders, he worked in LDI sales at ABN AMRO in London where he was responsible for developing and maintaining relationships and executing transaction with LDI fund managers. From 1998 to 2006 he was a Bond Fund Manager at Barclays Global Investors where his roles included Head of Multi-Currency Portfolio Management and Head of the Interest Rate Process. Investment career began in 1993 when he joined the financial business unit at KPMG. Andy has an MA in Land Economy from St. Catharine s College, Cambridge. He is a holder of the Securities Institute Diploma, Financial Derivatives and Bonds and Fixed Interest, and is a member of the Securities Institute and the Institute of Chartered Accountants. Mike Hodgson, Ph.D Head of Risk Managed Investment and Structuring Mike is Head of Risk Managed Investments and Structuring at Schroders, and is based in London. He rejoined Schroders plc in 2011. Mike has over 20 years experience in financial markets, having started his career in 1987 at J Henry Schroder & Co. Limited as Principal Interest Rate Derivatives Trader and then was promoted to Global Head of Structured Products and Interest Rate Derivatives. In 2000, he moved to Citigroup as a result of the purchase of J Henry Schroder where he was European Head of New Product Development until 2002 when he founded Hodgson Global. In 2004, Mike joined ABN AMRO Bank NV (which then became Royal Bank of Scotland NV in 2007) as Global Head of Equity Derivatives Structuring moving on to become Head of Fund Derivatives Trading and Structuring. Mike earned his Ph.D in Physics from Cambridge University and BSc (Hons) in Physics at Imperial College in London. Managing investment outcomes with volatility control 17
Important information: The views and opinions contained herein are those of the Schroders Portfolio Solutions Team, and do not necessarily represent Schroder Investment Management North America Inc. s house view. Issued December 2016. These views and opinions are subject to change. Sectors/regions/asset classes mentioned are for illustrative purposes only and should not be viewed as a recommendation to buy/sell. This material is intended to be for information purposes only and it is not intended as promotional material in any respect. The material is not intended as an offer or solicitation for the purchase or sale of any financial instrument mentioned in this commentary. The material is not intended to provide, and should not be relied on for accounting, legal or tax advice, or investment recommendations. Information herein has been obtained from sources we believe to be reliable but Schroder Investment Management North America Inc. (SIMNA) does not warrant its completeness or accuracy. No responsibility can be accepted for errors of facts obtained from third parties. Reliance should not be placed on the views and information in the document when taking individual investment and / or strategic decisions. Past performance is no guarantee of future results. The opinions stated in this document include some forecasted views. We believe that we are basing our expectations and beliefs on reasonable assumptions within the bounds of what we currently know. However, there is no guarantee that any forecasts or opinions will be realized. This document does not constitute an offer to sell or any solicitation of any offer to buy securities or any other instrument described in this document. All investments involve risks including the risk of possible loss of principal. The market value of the portfolio may decline as a result of a number of factors, including adverse economic and market conditions, prospects of stocks in the portfolio, changing interest rates, and real or perceived adverse competitive industry conditions. SIMNA Inc. is registered as an investment adviser with the U.S. Securities and Exchange Commission and as a Portfolio Manager with the securities regulatory authorities in Alberta, British Columbia, Manitoba, Nova Scotia, Ontario, Quebec and Saskatchewan. It provides asset management products and services to clients in the United States and Canada. Schroder Fund Advisors LLC ( SFA ) is a wholly-owned subsidiary of SIMNA Inc. and is registered as a limited purpose broker-dealer with the Financial Industry Regulatory Authority and as an Exempt Market Dealer with the securities regulatory authorities in Alberta, British Columbia, Manitoba, New Brunswick, Nova Scotia, Ontario, Quebec and Saskatchewan. SFA markets certain investment vehicles for which SIMNA Inc. is an investment adviser. SIMNA Inc. and SFA are indirect, wholly-owned subsidiaries of Schroders plc, a UK public company with shares listed on the London Stock Exchange. Further information about Schroders can be found at www.schroders.com/us or www.schroders.com/ca. Schroder Investment Management North America Inc. (212) 641-3800. WP-VOLCONTRL