Challenges and Opportunities for the year ahead as seen from the perspective of the AOFM – Presentation to the Australian Business Economists luncheon
13 September 2016
Presentation delivered by Rob Nicholl – CEO, AOFM
Thank you to the ABE for again offering the opportunity to present on our thoughts for the year ahead from an AOFM perspective.
As the title for the presentation today suggests, we are looking to the current year as a continuation of the risk management theme as well as being a year of providing opportunity. The opportunity as we see it will be to undertake further portfolio development and to introduce arrangements – such as bond maturity conversions – that will help us manage funding risk, while at the same time benefiting the market. I will provide more detail on these shortly. My focus will be on issues regarding nominal bonds because this comprises by far the largest part of our portfolio management task. However, I will provide some comments on linkers towards the end.
Looking back over the past few years I hope that the AOFM can be seen as having been clear in its objectives, while measured and confident in its approach. There was definitely a few years where we were less confident about, for example, how far we could reach in terms of yield curve extensions, or what would work as appropriate in-fill maturities. This was because forecasts at that time for future issuance needs and the likely outright size of the market were tempering our expectations on what developments would be sustainable. Once committing, it remains important to us to be able to support a new development over a meaningful forward period. What underlies this on our part is the presumption that this is useful for investors.
I made clear last year that we will continue to support the 20‑year part of the curve with new issuance, although we have found that to be more challenging at times than we had initially expected. In turn this will obviously deliver on our commitment to support the new futures contract.
While the new contract is not of our making, nor are the workings of it something over which we have any control, it remains our view that a successful 20-year futures contract will prove to be a useful market development.
Since this contract was launched I have spoken to many investors. The responses have been mixed – ranging from those having participated and hoping to be able to use it more frequently, to those who are sitting on the sidelines waiting for it to further develop. One thing we should all be aware of is that if a majority wait to see whether they can benefit from the actions of early users, then it will take a lot longer for everyone to benefit from the new contract. This is not a ‘paid commercial’ but a common sense prompt to consider that if all participants reserve just a little bit of what they do toward a broader market benefit then this can only be a good thing. While the AOFM should take a lead role in this I have seen examples of it in our market at times from other participants, so I know it is not only possible but a reasonable expectation.
Before I turn to specifics of the plans for this year I will give some context to what we see as relevant for the tasks ahead. Domestically we have forecasts for Budget financing requirements and in turn our issuance requirements that, while forecast to ease, will remain elevated nonetheless.
Last year we issued around $97 billion in term funding and this was achieved with the assistance of some sizeable syndications. Notwithstanding syndications, achieving this task still required a relatively high regular issuance rate. This year started with planned issuance of around $93 billion and so the remainder of the year is not going to be a markedly different task in that regard.
While commentary about the impact of Australia’s ongoing fiscal challenges is often discussed in terms of debt to GDP ratios – the required weekly issuance rate for any year is of most interest to us.
In this regard we are mindful of a number of external factors that can make it more difficult for us to issue at any point in time, or if they persist, over time. None of the resulting impacts or influences will, we hope, surprise us and a good part of what we do aims to keep our thinking ahead of changes in the market. With that in mind, the main issues we continue to monitor closely include:
- the growing impact on our intermediaries of regulatory changes and the ability and/or willingness of banks to allocate balance sheet to market making – and in particular to primary market participation;
- the volume of bonds sitting on trading books (which of course can be because of a slowing in demand and/or periods of net selling);
- movements in cross-currency levels, which have particular relevance to offshore investor appetite;
- the time of the year (which takes into account end-of-financial banking years, public holidays, major planned market related announcements, and so on); and
- yields, which of course do not always accurately reflect the level of point‑in‑time demand, but do impact the relative attractiveness of Australian Government bonds relative to substitute assets.
A number of these issues relate directly to price (or yield). While we see AOFM as very much a price taker in primary market auctions, we are aware of our ability to move bond prices through announcements and actions, although this is not an end we have ever set out deliberately to achieve. Periods of overly tight congestion weigh as heavily, if not probably more so, on our minds as they would for anyone else in the market. This is an important reason to conduct our activities so far as possible in a manner that promotes efficiency in the market – that is to everyone’s benefit, not just ours.
Turning to the other issues we are monitoring closely, these include:
- the status of and changes to monetary policies (and in particular the possibility of either extended, or alternatively reduced, quantitative easing in the major economies);
- the potential for even lower and more widespread negative yields in other sovereign markets, or conversely a sharp back-up in yields;
- the possibility for more strident and open currency competition in response to ongoing subdued economic outlooks;
- further development of China’s capital market;
- the compression in yield spread between our bonds and US Treasuries;
- appreciably reduced new allocations into the AUD and Australian Government Bonds by reserve managers; and
- geopolitical uncertainties that could create volatility and associated major financial market dislocation for any number of reasons.
In broad terms a mix of domestic and international issues condition our thinking on the downside risks to issuance programs and combinations of these have the potential to result in higher than forecast issuance, an appreciable selling of our bonds or a sharp decrease in demand for them. At the same time circumstances offer opportunities to further develop the long-end of the yield curve (which I take to include 10-year futures stocks and longer) and to issue into available demand by taking account of prevailing market conditions.
The following are some initiatives that aim to cover these challenges and opportunities.
I mentioned that this year started with a relatively high issuance program. Bearing this in mind, an important part of our role is to plan for and create the flexibility that would be required to manage a material increase in the financing task, regardless of the cause. There are various means at our disposal for this. In past years, for example, we have issued disproportionately more into the first half of the year to get ahead in the program prior to the annual summer shut-down. Increasingly elevated annual issuance tasks can reduce that flexibility.
Also in the past the number, timing and nature of new maturity syndications has been used as a means of balancing market development with an opportunity to achieve issuance volume.
Very large maturing bond lines is also something we have to consider given the volume of issuance that has been required over the past five years or so. This means that the starting point for annual issuance is increasing over time independently of changes to financing tasks required to meet Budget deficits.
But a commitment to support the three-year futures contract and to provide for investors who are focused on the short – to – medium part of the curve means that we will continue to allocate issuance into three to seven year maturities. To some extent, the within-year funding risk associated with large bond maturities has been reduced by introducing a second calendar year bond line – something we started a few years back and have extended to the 11-12 year part of the yield curve. This is now considered to be established operational practice.
We have reduced the annual funding risk by extending the yield curve and in part by allocating increasing volumes of issuance into the long-end.
This chart shows the proportion of issuance into various parts of, and lengthening to, the yield curve that have given effect to this strategy. Looking to the right hand side – the light bar at the top represents the three-year futures basket (indicating no issuance into bonds shorter than this since 2013-14); with the mid-bar representing the 10-year futures basket.
In 2011-12 we lengthened the yield curve to around 15 years with just less than 40 per cent of issuance into 10-year futures stocks and beyond. By July this year, the curve had been extended to around 23 years with just over 60 percent of issuance into 10-year futures stocks and longer. As a result, annual funding risk over the next five years has already been reduced by about $12 billion per year. In order to continue with this strategy, the intended issuance pattern for this year will be broadly similar to that of the past two years.
A new mechanism to help manage the annual funding risk will be bond maturity conversions. In May we foreshadowed plans to convert short-dated Treasury Bonds into longer Treasury Bond lines. This will be achieved through buyback tenders that are funded by additional issuance, with the first of these to be held next Monday. This will be additional issuance to the $93 billion program we announced at the time of the Budget.
The early buybacks of short-term bonds will offer a number of benefits. They will increase portfolio duration, and assist in the cash management task at bond maturities. They will also enhance the operation of the secondary market, by removing stock left on trading books as short‑end bonds drop out of indices or the three-year futures contract. Finally, they will assist the RBA in its liquidity management task, which will otherwise become increasingly challenging over the years ahead on days of bond maturities.
There are three ways in which we can repurchase bonds prior to maturity: one is the buyback tenders; another is through offering switches as part of future syndicated issues; and a third is by buying back bonds the RBA has purchased through its liquidity operations.
Because AOFM buying bonds back from the RBA has no additional market impact, our intention is to complement rather than substitute for RBA liquidity operations by targeting bonds as they fall out of the three‑year futures basket and up until they are around 12-18 months out from maturity.
We are aware that a surplus of short bonds doesn’t necessarily equate to there being no investor interest in them, so this process will begin conservatively and feedback on liquidity will be monitored as the process builds over time. However, very short-end liquidity will remain more a consideration than a determinant of the rate at which maturity conversions will build.
I will now turn to the long-end of the curve, which is where we tend to see opportunity.
The AOFM has for a number of reasons been cautious to date in its market development role. Incremental yield curve extensions are based primarily on portfolio considerations. Once it became clear that issuance programs were going to be elevated for some time, this gave yield curve extensions both prudency and legitimacy. Prudency because we have been able to reduce the portfolio impact of potential future interest rate volatility, with the added benefit of reducing the funding risk I highlighted earlier. Legitimacy because we were confident that an extended yield curve could be supported over the long-run rather than seeing new investors with an appetite for duration attracted to the market, only to find that they were unable to build and manage portfolio allocations based on reliable supply.
The following two charts illustrate the effects of the lengthening strategy.
This chart shows the average term to maturity of the nominal bond portfolio coming immediately out of the global financial crisis. It highlights the impact of abruptly higher issuance on the average term to maturity as a sharp increase in the volume of short bonds occurred, but it has since been rising steadily and is currently around seven years. At the same time the average cost of the outstanding stock of debt declined by around 160 bps over the same period.
This chart shows the marginal impacts on average term to maturity and cost against new issuance each year. It is again clear from this chart that immediately following the financial crisis the average term to maturity of new issuance declined, then increased markedly to a peak in 2014-15, which is the year we established the 20-year part of the yield curve. At the same time the average cost of new issuance has declined by over 270 bps.
Whether or not we should be ‘locking in low rates’ by issuing into the long-end is not the prime perspective from which we view portfolio development. That the Government can now borrow at lower cost than it could several years ago is a consequence of global economic and financial market influences, it does not in itself reflect an AOFM objective. To us it is more a convenient coincidence that the context in which we are operating has allowed extending the average term to maturity of the portfolio (which includes having extended the yield curve from 12 to 24 years) at lower than historical cost. That we have been able to do this comes down to the fact that rates are low globally because of the extraordinary attempts by central monetary authorities to stimulate economic growth through various channels. The good news out of all of this we are yet, of course, to discover.
That said, development of the long-end will continue, including as I noted with weighting at least as much issuance into these maturities as we achieved in the past two years. Trying to determine how hard we should push this objective is debatable, although I remain of the view that issuing regularly (even if not frequently) into the long-end will be beneficial.
Supporting the 20-year part of the curve is a key consideration for reasons I have already highlighted and establishing a new 2041 maturity is an obvious part of that aim. This is because we intend to maintain two-year maturity gaps in this part of the curve. Having established a new maturity last year you can expect us to repeat this pattern with the 2041 maturity which can be expected around this time next year.
In the meantime we are also planning to assist liquidity in this part of the curve by the use of small syndications into existing 20-year benchmark maturities from time-to-time. While we are still refining our thinking as to how these might be most effective, if successful we would look to use them as an ongoing measure (maybe up to several times a year), as a complement to smaller tender volumes as and when this seems appropriate. As is always the case, regular market feedback will help to determine sensible timing and volume for both the syndications and the tenders. However, we are not contemplating the first of these syndications until sometime after February next calendar year.
This leaves the issue of a further yield curve extension.
It will probably not come as a surprise that we would see the completion of our market development activities resting with the establishment of a 30-year bond, although we have to date been reluctant to commit to a firm issuance plan. But we are now of the view that the time is right for this to proceed. We are planning, subject to market conditions, to do this around the second week of October. While I am not going to share the specifics today, a detailed notice will be provided in the usual manner prior to launching the deal. What I will clarify though is that the term ‘subject to market conditions’ refers to our usual consideration that there is no expectation of market volatility during the process that would create uncertainty and unnecessary challenges at the time of pricing the new bond.
Someone is bound to ask how much we are expecting to issue and for other detail such as expected pricing. As to the first, let me say only that we are expecting to be able to issue a meaningful volume to meet anticipated broad interest and to establish a maturity of clear benchmark size. We are confident that domestic investor support and the so-called ‘global search for yield’ that has underpinned an increased demand for duration will remain sufficient for us to establish a 30-year Australian yield curve.
While we have been thinking about pricing for a while, that is not something I am prepared to share and see it more as one of several deal specific issues to be discussed with the banks that are to be invited as lead managers for the deal. How often this will be repeated is an open question. At present there seems to us to be two obvious choices of pattern to follow; one is a new 30-year issue every four years, or in the event that sufficiently strong interest builds for these maturities another option may be to consider repeating a new issue every second year. However, this is not something that needs to be decided now and we will be open to advice as to what is appropriate in this regard.
We are mindful that speculation around our plans for new maturities by syndication can create both positive and negative tension, but in this case we are taking the view that ‘constructive ambiguity’ (as the outgoing RBA Governor would say) is not going to be helpful. In this regard I want to reiterate that a new 2041 maturity is not planned until at least around this time next year. Furthermore, the new 2028 maturity to support the 10-year futures contract will not be launched this calendar year. And, as I also said earlier, we will not be trialing the smaller syndications of existing 20-year maturities until sometime after February next calendar year.
I have included the following chart as a summary of the issuance related detail provided today. It shows the region to be targeted for maturity conversion buybacks along with the regions of the futures contracts we will continue to support through issuance and the new lines to be established through syndications this financial year.
This leaves me with several other things to comment on by way of an update. One is the linker market and our announced program for this year; and another is RMBS.
On linker issuance, we have responded to subdued demand by announcing a cautious program for the year ahead. We recently conducted a comprehensive schedule of meetings with domestic investors and it is fair to say that we expect our intention to treat the market carefully will be appropriately received. Feedback on liquidity indicates it is, as some investors have described, ‘challenging‘. What further role the AOFM could or should play in providing liquidity is an open question and we will continue to consider a range of options on this, together with their associated merits and costs.
A commitment to maintain the market remains firmly in place, but how that is achieved in the current context is something that remains less obvious than is the case with the nominal bond market. That said we have a lot of operational flexibility to pursue this objective. I use the term ‘operational flexibility’ cautiously because exercising flexibility in the past has often been interpreted as opportunism. I have repeatedly sought to quell this perception, but to quote the outgoing RBA Governor – just because our actions could be interpreted the wrong way doesn’t mean we should hold back from doing what we think is the right thing to do!
Consistent with our May announcement there are no near‑term plans to extend the linker yield curve and the possibility of establishing a new 10‑year maturity remains an open consideration, but not before early 2017. Maintaining the current pattern of modest volume tenders and taking into account regular market feedback, therefore, appears to us to remain a prudent course to follow.
Turning to RMBS, we remain heavily reliant on market conditions generally as to when to resume the divestment process. Four auctions were held prior to the deteriorating market conditions that gave us cause to suspend the process. Given that the Government is not a distressed seller of the remaining portfolio, the AOFM is now focused on criteria that weigh holding a low risk portfolio against selling RMBS at a capital loss to the taxpayer.
While we have no better view than anyone else as to when conditions might improve we continue to monitor the market both in terms of new primary issuance and other relevant information. For example we follow the markets for comparable instruments, such as domestic senior unsecured bank paper. We note that this market lagged the offshore market late last calendar year as spreads were widening (with RMBS lagging still further behind), which could be interpreted as the offshore market leading the domestic market. It would appear that the offshore market for bank paper is again leading the domestic market as spreads contract. If this were to continue, it would suggest that RMBS is currently undervalued, but this will be a judgment best made by those closer to the action.
I will conclude by offering some higher level observations on the Australian sovereign bond market. From our perspective the market remains resilient and relatively efficient although we note periods of congestion. Investor feedback on liquidity, however, continues to be positive overall.
The proportion of offshore ownership has been steadily declining and is currently around 60 per cent, compared with a peak of around 78 per cent a few years ago. It is difficult to determine in detail what lies behind this trend but we do know that:
- reserve manager accumulation of Australian Government Bonds is a mature rather than a maturing story;
- overall, net offshore buying continues steadily – although we do experience periods of notable net selling – but the rate at which offshore buying has increased over the past few years has not met the rate of increase in our issuance; and
- the market continues to attract new investors with a lengthening yield curve playing a notable role in this further investor base diversification.
The declining significance of offshore investors is not in itself of particular concern as we look to our current and forecast issuance tasks, although it does remind us that there are likely to be limits to offshore demand for our bonds in AUD and at current yields relative to alternative assets.
We remain confident that there is sufficient diversity in issuance options and in the investor base for the market to continue to operate efficiently, and we live in the hope of course that major external factors will not test our thinking on these matters – although on this we cannot be sure.
Last updated: 14 September 2016