DIRECT INFRASTRUCTURE VALUATIONS AND BOND RATE INCREASES:

Transcription

1 insightpaper DIRECT INFRASTRUCTURE VALUATIONS AND BOND RATE INCREASES: it s not what you expect April 2017 AMP CAPITAL INFRASTRUCTURE 1

2 Key points Future bond rate increases are likely to be moderate. AMP Capital s Investment Strategy and Economics team (house view) forecast an increase in Australian 10-year government bond rates from approximately 2.7% currently (27 March 2017) to approximately 4% in three years. > > Rising bond rates will tend to have a negative impact on unlisted infrastructure valuations, but rising rates should not be considered in isolation, as this is only one of a number of factors which influence valuations. > > We believe that current unlisted infrastructure valuations contain significant buffers that may act to offset the impact of bond rate increases. Current buffer margins could accommodate cumulative bond rate increases of up to 1.5%. > > Additionally, the pressure for future bond rate increases is coming from improved economic conditions. This same pressure is likely to have a positive impact on cash flows, particularly for GDP- linked growth assets. This is likely to have a positive impact on valuations. > > The net impact of bond rate movements will therefore vary among different infrastructure classes. > > Our analysis considers net impacts on three classes of infrastructure assets: GDP-linked growth assets; Bond-proxy PPP concessions, and; Regulated utilities. > > Our analysis shows that: Growth infrastructure asset valuations can still grow strongly in the face of moderate bond rate increases, but larger increases could restrict growth (although total returns are likely to remain positive). Bond-proxy PPP concession asset cash flows may benefit indirectly from a rate increase and may also be quite resilient to modest bond rate movements. Consequently, these assets should outperform traditional bonds in a rising rate environment. We expect the strong yield characteristics of these assets means they should play an important role within a well-diversified infrastructure portfolio, particularly where investors investment objectives include a healthy level of income. Regulated utility valuations are largely indifferent to bond rate movements. > > While the analysis is based on Australian infrastructure, we would expect that broadly similar results would be found by similar analysis of foreign assets. 2 AMP CAPITAL INFRASTRUCTURE

3 Introduction Since the global financial crisis, government bond yields in developed economies have been driven down by weak economic growth and excess liquidity. This downward momentum appears to have halted following recent improvements in sentiment and economic conditions in the United States (US), Japan and Europe. The improving economic backdrop in the US has been confirmed with the US Federal Reserve raising official rates by 0.25% in March US forecasts are suggesting real economic growth over the next 2 years (to the end of 2018) 1 will reach a little north of 2%. This is encouraging but falls short of the good old days of +4% GDP growth per annum. This suggests that any accompanying inflationary pressures can be managed with moderate government bond rate increases. In Australia, AMP Capital s house view supports the case for moderate government bond rate increases, as summarised in figure 1. Figure 1: Actual and forecast returns for cash rates and bond yields FORECAST PERIOD AUSTRALIAN CASH RATE AUSTRALIAN 10 YR. GOVT. BOND RATE Current 1.50% 2.70%* 1 year 1.25% 3.50% 3 year 2.00% 4.00% 5 year 3.00% 4.5% Source: AMP Capital as at February 2017 * As at 27 March 2017 Over the longer term, AMP Capital s Global Fixed Income team anticipate that increases in developed market bond rates will be constrained by secular influences, such as aging population demographics, which will reduce potential economic growth. As a result, yields over the coming decade or so are expected to be lower than the average of the twenty years preceding the GFC. In our view, medium and longer term bond rate increases will be moderate. 1 US Federal Reserve Bank AMP CAPITAL INFRASTRUCTURE 3

4 How do changes in bond rates impact unlisted infrastructure? AMP Capital s Global Listed Infrastructure team recently published a paper, Global Listed Infrastructure: Not just a bond proxy that examined the impact of rising bond yields on listed infrastructure market caps. The analysis focused on the significant sell off of infrastructure stocks following the US Federal Reserve s advisory note, in July 2016, which flagged three potential increases in the cash rate out to the end of The question arises as to what impact bond rate increases might have on unlisted infrastructure asset valuations. In contrast to listed infrastructure stocks, unlisted infrastructure valuations are largely immune to listed market sentiment, but valuations can still be impacted by rising bond yields through a number of mechanisms: 1. Higher interest rates may impact operational cash flows Increasing government bond yields will pass through into interest rates which in turn may impact operational cash flows through higher debt servicing costs. Consequently, the management of an asset s debt portfolio is a primary focus of risk control. Asset managers employ a range of sophisticated risk management techniques, including: > > Diversifying debt portfolios by both provider and tenor. This allows the manager to take advantage of favourable market conditions; > > Using interest rate swaps to hedge the risk of future underlying bond rate movements; > > Issuing long dated bonds in foreign markets, hedged to Australian currency. Where asset gearing is conservative and debt portfolio management is effective, the valuation impact of moderate interest movements on operational cash flows is relatively minor. Consequently, this paper focuses on impacts from the discount rate. 2. Higher government bond yields may impact the discount rates used in valuations Unlisted infrastructure valuations are regularly undertaken by the major international accounting firms, typically using a discounted cash flow analysis (DCF) of forecast future cash flows. The final valuation will be checked against comparable market transactions. Government bond yields are an important input, as they are often used as a proxy for risk free rates and impact both the future cost of debt and equity. Consequently, if taken in isolation, rising rates can be expected to negatively impact asset valuations by increasing the discount rate (see Box 1). However, looking at bond yields in isolation can be very misleading as most unlisted infrastructure asset valuations are also impacted by a number of factors, some of which may offset the pure bond yield impact. Box 1. Typical unlisted infrastructure valuation methodology Discount rates used in the discounted cash flow (DCF) analysis are calculated by estimating the weighted average cost (WACC) of debt and equity. When applied to an asset s forecast future cash flows, this allows the enterprise value of the asset to be calculated. The value of the equity component (which is commonly thought of as the asset value from an equity investor s perspective) is obtained by subtracting the value of debt from the enterprise value. For infrastructure assets, the cost of equity is typically estimated using the Capital Assets Pricing Model (CAPM), modified to incorporate asset specific risks, according to the following formula: Ke=Rf +β(rm-rf) + α where: Ke (or discount rate ) = required return on equity; Rf = the risk free rate, typically a valuer will use an averaged spot zero coupon yield on a 10 year government bond as a proxy for the risk free rate; Rm= the expected return on the market portfolio; β= equity beta, the systematic risk of a particular type of stock (e.g. regulated utilities); α= asset specific risk premium Effectively, the 10 year government bond rate directly sets the risk free component of the discount rate, while the β(rm-rf) and α elements of the equation determine overall equity risk margins. These are discussed on the next page. 4 AMP CAPITAL INFRASTRUCTURE

5 Factors that can mitigate bond yield movements Valuation buffers Stronger cash flows Since the GFC, valuers have been reluctant to fully pass through the impact of the bond rate reductions that followed. Consequently, risk margins increased over this period and are now at record high levels, being between 1.2% and 1.5% (depending on asset type) higher than those applied to similar assets prior to the GFC. In part, the high risk margins reflect valuers concerns about the sustainability of low bond yields in the valuation of longdated assets. The prospect of improved economic conditions is the basic driver of the anticipated bond rate increases. Growth-linked infrastructure asset cash flows, in particular, are leveraged to economic growth, so we would expect improved growth to be reflected in stronger cash flows and have a positive impact on valuations. As government bond rates increase, we therefore expect valuers will unwind the high risk margins, although the timing is uncertain. The main point is that the high risk margins in current unlisted infrastructure valuations provide a significant buffer against future long-term government bond rate increases. A number of different mechanisms mean that cash flows for PPP assets may also be stronger during periods when interest rates are rising (refer Case study 2 below), with a positive impact on valuations. Therefore, we anticipate that the net effect of bond rate movements on valuations will be the sum of: >> the positive impact of unwinding of risk buffer impacts on discount rates; >> The positive NPV impact of stronger cash flows; >> The negative impact of bond rate changes on discount rates. CASE STUDIES We have examined the potential valuation impact of moderate government bond rate increases on three classes of infrastructure assets, through case studies on: 1. GDP-linked growth assets. Typical assets include ports, airports and toll roads where the operator is exposed to price and demand risk. The sensitivity of equity valuations is modelled against a number of scenarios. 2. Bond proxy (high yield) assets. Bond proxy assets primarily involve availability payment-based Public Private Partnership concessions (PPPs). The sensitivity of equity valuations is modelled against a number of scenarios. 3. Australian regulated utilities. Typical examples include energy transmission and distribution companies. Given the low sensitivity of this asset class to bond rate movements, the sensitivity of equity valuations is not modelled. The case studies are based on AMP Capital modelling of generic asset types rather than actual assets, using elasticities of valuations to risk factors which are typical of Australian assets in the asset class. A two year period is modelled. AMP CAPITAL INFRASTRUCTURE 5

6 Case study 1: Growth (GDP-linked) asset In growth assets, returns are predominately generated through capital growth. Operational risk includes price and patronage risk and is substantially higher than for regulated utilities or bond proxy assets. Consequently, the required cost of equity (that is, the required equity return to investors) is higher. The different scenarios modelled and assumptions used are set out in Box 2. To assist in visualising the various impacts, Figure 2 depicts a two year valuation bridge, representing Scenario 2. For simplicity, the assumed valuation period is 12 months, rather than the normal 6 months. Actual modelled values for all three scenarios are contained in Appendix A. Box 2. Growth asset valuation scenarios SCENARIO 1: Base case, no change in rates - bond rates and risk premium buffer remain unaltered at 2.5% and 1.5% respectively. SCENARIO 2: Government 10 year bond rates rise to 4.0% over 2 years. They increase by 1.00% per annum in Year 1 and 0.5% in Year 2. The risk premium buffer is utilised at a rate of 75% of the bond rate movement, leaving 0.38%of the buffer unutilised. SCENARIO 3: Government 10 year bond rates rise to 5.0%, increasing by 1.25% per annum in each year. The risk premium buffer is fully utilised by the end of Year 2 in two equal tranches. Assumptions Base case returns: 10% (total return) pa; 4% pa yield Revenue/Nom. GDP ratio: 2.5:1 Elasticity Equity Value/Discount rate: -9:1* Start rate for Australian 10 year government bonds: 2.5% *correlation from actual assets Figure 2: Two year valuation bridge for a growth asset under scenario Previous Valuation Time value of money Delta Cash flows Asset Risk Premium Movement Bond Rate Movement Dividends Y1 Valuation Time value of money Delta Cash flows Asset Risk Premium Movement Bond Rate Movement Dividends Y2 Valuation Source: AMP Capital 6 AMP CAPITAL INFRASTRUCTURE

7 Explanation of each valuation driver: 1. Valuation roll forward (or time value of money) - This represents the impact of moving forward the valuation date, in this case by 12 months at a time. Consistent positive growth in long-dated assets will produce a value uplift as the starting date of the DCF analysis moves forward on progressively increasing cash flows. The higher the growth rate of future cash flows, the higher will be the value lift. The roll forward in this example assumes no change from the base case as cash flow and discount rate impacts are captured in the following items. 2. Delta Cash Flows - Growth assets benefit from economic growth, so we can expect that increased cash flows from improving economic conditions will flow to the bottom line, as operating costs are largely fixed. 3. Asset Risk Premium Movement - Assumed decreases in risk margins of 0.75% in year 1 and 0.38% in year 2, leads to respective increases in equity value of 6.75% and 3.38% (using an Equity Value/Discount rate elasticity of -9:1). 4. Bond Rate Movement - Using the same elasticity, the assumed 1% increase in bond rate drives a 9% reduction in equity value in year 1 and the assumed 0.5% increase in year 2 drives a 4.5% reduction. 5. Dividends - A base case payout of 4% is maintained. The results for all 3 scenarios are summarised in Figure 3. Figure 3: Growth asset scenario analysis: valuation impact of bond rate change Relative Equity Value Valuation Y0 Valuation Y1 Valuation Y2 Scenario 1 - Base Case Scenario 2- Rise to 4% Scenario 3 - Rise to 5% Source: AMP Capital Results of case study 1: growth asset scenario analysis SCENARIO 1 Base case, no change in Government 10 year bond rates. Valuations rise at 6% per annum which gives a total return of 10% per annum when dividends are included. This is consistent with a moderate performing growth asset. SCENARIO 2 Government 10 year bond rates rise to 4%. Valuations outperform the base case over 2 years, despite the incomplete wind-back of risk margins. Returns average 10.4% including a 4% dividend. SCENARIO 3 Government 10 year bond rates rise to 5%. In this downside scenario, valuations underperform the base case. Returns average 8.7% including a 4% dividend. The results suggest that growth assets valuations can be highly resilient to moderate bond rate movements as a result of both valuation buffers and their exposure to improving economic conditions. Further analysis suggests that government bond rates would have to rise to about 6.5% before overall valuations show a negative change, and 7.2% before overall returns become negative. AMP CAPITAL INFRASTRUCTURE 7

8 Case study 2: Bond proxy PPP concession assets Box 3. Bond proxy assets valuation scenarios SCENARIO 1: Base case, no change in rates - bond rates and risk premia buffer remain unaltered at 2.5% and 1.25%, respectively. SCENARIO 2: Government 10 year bond rates rise to 4.0% over 2 years. They increase by 1.00% per annum in Year 1 and 0.5% in Year 2. The risk premium buffer is utilised at a rate of 60% of the bond rate movement leaving 0.38% unutilised. SCENARIO 3: Government 10 year bond rates rise to 5.0%, increasing by 1.25% per annum in each year. The risk premium buffer is fully utilised by the end of Year 2, in two equal tranches. Assumptions Base case returns: 9% (TR) pa; 9.5% pa yield Equity Value/Discount rate: -8:1 Start rate for Australian 10 year government bonds: 2.5% High yielding assets have been in strong demand as investors seek bond proxies. Availability-type PPP concession type assets have been especially attractive given the absence of demand risk and a strong counterparty (government). Typical features of these assets are: 1. They fully amortise over the concession period (typically 25 years). Consequently: a. valuations will decline over the period and roll forward may be negative; b. returns will include both return on capital and return of capital, meaning that the yield may be higher than the return on capital. 2. In comparison to high growth assets, risk margins are significantly less, reflecting the secure nature of PPP cash flows. The lower risk means that: a. assets are more highly geared than typical growth assets. Consequently, the trade-off between lower risk margins and higher gearing results in an elasticity of equity valuations to discount rate of -8:1 2. b. valuation risk buffers are lower as a result of the lower equity beta. 1.25% is assumed. 3. Day to day operations may be contracted out to specialist operators, who may also accept asset refurbishment risk. High yielding assets have been in strong demand as investors seek bond proxies. Availability-type PPP concession type assets have been especially attractive 4. The major operating risk remaining with the investor, relates to the periodic need to refinance debt during the concession period; 5. Cash flows may be adjusted for inflation but are not otherwise impacted by changes in economic conditions. 6. Operational improvement benefits may be shared with the external operator. 7. With limited capital requirements, all free cash is paid as dividends. The low risk and strong government counter party mean that these assets are often viewed as bond proxies. Valuers may take a slightly different approach when valuing bond proxy assets, compared to valuing growth assets, replacing the market risk premium in the above CAPM equation (refer Box 1) with a credit premium, giving: Ke=Rf + Cm + α where: Cm = credit margin with other terms remaining as above. 2 AMP Capital modelling 8 AMP CAPITAL INFRASTRUCTURE

9 In addition to the valuation risk buffer, PPP assets may also have a number of further protections against bond rate increases: 1. Pain/Gain share arrangements are common, where the debt margin risk is shared with the counterparty. For simplicity, a pain/gain share buffer has not been modelled. 2. A full or partial pass through of CPI provides a degree of hedging against underlying bond rate movements, because CPI could be expected to rise with bond rates. The Reserve Bank of Australia expects inflation pressures to be modest in the medium term, with a possible rise of 0.5% pa over the current level of 1.5% over the next two years 3. The impacts of a 75% CPI pass- through of this increase 4 result in a strengthening of cash flows. 3. Lenders require high cash reserves to be held in order to meet debt service cover ratios. These reserves can exceed 30% of the equity value of the asset 5. Cash rates tend to move in concert with long term government bond rates 6. Increases in cash rates will flow through as increased interest returns on these cash reserves, providing a significant offset to the bond rate impact. This further strengthening of cash flows results in a positive impact on equity valuations. The results are summarised in Figure 4. Associated data is contained in Appendix B. Figure 4: Bond proxy asset scenario analysis: valuation impact of bond rate change Relative Equity Value Valuation Y0 Valuation Y1 Valuation Y2 BaseCase Scenario 2- Rise to 4% Scenario 3 - Rise to 5% Source: AMP Capital 3 RBA Statement on Monetary Policy Feb Source: Based on an average of a portfolio of 12 Australian PPP assets held in the AMP Capital Community Infrastructure Fund 5 Source: Based on an average of a portfolio of 12 Australian PPP assets held in the AMP Capital Community Infrastructure Fund 6 RBA Data since June 1976 AMP CAPITAL INFRASTRUCTURE 9

10 Results of case study 2: bond proxy PPP asset scenario analysis SCENARIO 1 Base case, no change in Government 10 year bond rates. Valuations decline over time as a consequence of the amortising nature of these assets. Return on equity is about 9.0% including a 9.5% dividend. SCENARIO 2 Government 10 year bond rates rise to 4%. The initial decline in valuations was significantly greater than for the base case, but recovered strongly in Year 2 as a result of the cumulative impact of stronger cash flows and the valuation buffer. Dividends were maintained at 9.5% and total return on equity averaged 9.8% over the 2 year period. The additional return would eventually be manifest as higher yields. SCENARIO 3 Government 10 year bond rates rise to 5%. Valuations declined by about 1% more than the base case over the two years. Dividends were maintained, but total return on equity dropped to approximately 8.5%. The results suggest that, although the mechanisms differ from growth assets, bond proxy asset valuations are also resilient to bond rate movements. While not having direct exposure to improving economic conditions, they benefit significantly from the pass through of any associated increase in inflation and interest returns from cash deposits. The results suggest that, although the mechanisms differ from growth assets, bond proxy asset valuations are also resilient to bond rate movements. 10 AMP CAPITAL INFRASTRUCTURE

11 Case study 3: Regulated Utilities As the name implies, regulated utilities are subject to economic regulation by a relevant regulator. Service standards, levels of operating and capital cost recovery, and returns to shareholders are reset on a regular basis (normally 5 years). The regulatory objective is to set a rate of return that delivers sufficient incentives to support investment in safe and reliable networks. Our approach allows us to determine a rate of return that is commensurate with efficient costs, reflects market conditions and is in the long-term interests of consumers - Australian Energy Regulator 7. The regulatory process tends to regulate revenue, so demand risk is low. However, under Australian regulatory conditions, the asset base, which provides a proxy for value growth, can be broadly expected to grow in line with long-term GDP. A modified CAPM is used by the regulators to determine returns on equity. Traditionally, long-term average bond rates were used in setting the risk free rate (Rf). However, this could lead to cycles of shortfalls and windfalls depending where the average rate sat relative to current bond rates resulting in uneven investment in assets. Recent trends in Australian energy utility regulation include: > > A move to use spot bond yields (averaged over 20 days) in the risk free rate determination. This means that broad bond rate movement impacts on equity returns are captured by the cost of capital reset. > > Changes in the cost of debt may be passed through to the operator on an annual basis. The result of these changes is that the net impact of bond rate movement on valuations is less than either growth or bond proxy assets. The resulting low risk means that the utilities sector suits investors seeking yield and moderate capital growth and moderate total returns over the medium to longer term. The resulting low risk means that the utilities sector suits investors seeking yield and moderate capital growth and moderate total returns over the medium to longer term. It should be noted that these comments apply to Australian energy utilities where the regulatory framework is specifically designed to minimise the impact of bond rate movements. Regulatory approaches vary significantly across jurisdictions, which could result in different sensitivities to bond rate movements. 7 AER rate of return fact sheet April 2015 AMP CAPITAL INFRASTRUCTURE 11

12 In summary The examples given in this paper are designed to illustrate the various ways in which bond rate movements impact unlisted infrastructure valuations. The scenarios modelled are based on our house view, together with a downside case. The results are illustrative and individual actual asset valuations may differ in their specific response to bond rate increases. However, the results should be broadly similar for each class of asset. > > Overall, with improving economic conditions and an associated moderate increases in bond rates, unlisted growth asset valuations may actually increase, relative to a scenario where there is no bond rate increase. This should not be a surprise, as growth asset returns are, by definition, leveraged to economic activity. > > Additionally, high yielding bond proxy PPP assets can prove to be resilient to moderate bond rate increases. In this regard, we would expect these types of assets to perform better than conventional bonds in a rising interest rate environment > > While not specifically modelled here, utility type assets are relatively indifferent to bond rate movements. The regulatory controls, implicit in utility type assets, provide protection against increasing bond rates for a long-term investor, albeit at the risk of some possible short term volatility depending on where the asset is situated in the regulatory cycle. > > While the analysis presented focuses on Australian assets, similar analysis can be applied to infrastructure assets in different jurisdictions. We would expect: Growth assets to show similar results to those presented above; Bond proxy PPP s may show some different characteristics. For example, while the impact of bond rate movements on valuations may be similar, such assets in the UK and Europe are often funded by long duration debt that matches the concession term, which eliminates the refinancing risk and hence the risk to operating cashflows. Yields also tend to be markedly poorer than Australian assets, making these assets less attractive on a risk adjusted basis, in our opinion; Foreign regulated utilities may be subject to quite different regulatory arrangements. However, underpinning all such assets is the need to maintain service quality by providing sufficient returns to attract investment. Consequently, we expect asset valuations, in mature regulatory environments, to demonstrate a similar low impact from bond rate increases. > > The timing of the wind back of risk margin buffers may lag bond rate movements. This may lead to some short-term volatility in valuations, which should even out over time. 12 AMP CAPITAL INFRASTRUCTURE

13 Appendix A Australian Growth Asset 2 Year Valuation Bridge Data SC 1 - BASE CASE SC 2- RISE TO 4% SC 3 - RISE TO 5% VALUATION Y VALUATION ROLL FORWARD DELTA CASH FLOWS ASSET RISK PREMIUM MOVEMENT BOND RATE MOVEMENT - (9.0) (13.5) DIVIDENDS (4.0) (4.0) (4.0) VALUATION Y TIME VALUE OF MONEY DELTA CASH FLOWS ASSET RISK PREMIUM MOVEMENT BOND RATE MOVEMENT - (4.7) (11.6) DIVIDENDS (4.2) (4.2) (3.9) VALUATION Y AMP CAPITAL INFRASTRUCTURE 13

14 Appendix B Australian Bond Proxy Asset 2 Year Valuation Bridge Data SC 1 - BASE CASE SC 2 - RISE TO 4% SC 3 - RISE TO 5% VALUATION Y TIME VALUE OF MONEY (1.0) (1.0) (1.0) CASH FLOWS AFTER EXPENSES ASSET RISK PREMIUM MOVEMENT BOND RATE MOVEMENT - (8.0) (10.0) DIVIDENDS (9.5) (9.5) (9.5) VALUATION Y TIME VALUE OF MONEY (1.0) (1.0) (1.0) CASH FLOWS ASSET RISK PREMIUM MOVEMENT BOND RATE MOVEMENT - (3.9) (9.8) DIVIDENDS (9.5) (9.5) (9.5) VALUATION Y AMP CAPITAL INFRASTRUCTURE

15 Overall, with improving economic conditions and an associated moderate increases in bond rates, unlisted growth asset valuations may actually increase, relative to a scenario where there is no bond rate increase. AUTHOR Greg Maclean Head of Research, Infrastructure AMP CAPITAL INFRASTRUCTURE 15

16 CONTACT DETAILS For more information on how AMP Capital can help grow your portfolio, visit our website, Important note: While every care has been taken in the preparation of this document, AMP Capital Investors Limited (ABN , AFSL ) makes no representation or warranty as to the accuracy or completeness of any statement in it including, without limitation, any forecasts. Past performance is not a reliable indicator of future performance. This document has been prepared for the purpose of providing general information, without taking account of any particular investor s objectives, financial situation or needs. An investor should, before making any investment decisions, consider the appropriateness of the information in this document, and seek professional advice, having regard to the investor s objectives, financial situation and needs. This document is solely for the use of the party to whom it is provided and must not be provided to any other person or entity without the express written consent of AMP Capital. Copyright 2017 AMP Capital Investors Limited. All rights reserved.