dairy: liquid powder - asb bank · • nzx wmp futures meet most of the criteria for the design of...
TRANSCRIPT
NZX Whole Milk Powder Futures 7 December 2015
Tobin Gorey – Agri Commodities Strategist – 612 9117 1130 – [email protected] Madeleine Donlan Agri – Commodities Analyst – 612 9303 8054 – [email protected] Jarred Rudman – Graduate - 612 9117 1130 [email protected] Nathan Penny – Rural Economist – 649 448 8778 – [email protected] Please refer to the important disclosures at the end of this document
Dairy: Liquid Powder NZX Whole Milk Powder (WMP) futures have taken some big developmental steps since their inception in 2010. Now the industry needs to put its mind to how the futures market can continue build liquidity and become the pricing benchmark for dairy products more generally.
EXECUTIVE SUMMARY
• The global dairy market has expanded rapidly in the past decade as many more people around the world enjoy the more, and better, food they can now afford. New Zealand dairy’s growth has been instrumental in this process as a larger cross-border flow of dairy, especially in Whole Milk Powder (WMP), has evolved to facilitate the demand from new customers. A new market needs new infrastructure to support that. Global Dairy Trade has facilitated the development of the physical market. Price volatility is generating a demand for hedging products that, in turn, would require a liquid derivative market. NZX’s WMP futures are the strong candidate to fulfil that role.
• NZX WMP futures have made the most progress among the dairy futures towards becoming a liquid benchmark. Trading volume and open interest have grown substantially and have developed in important ways that promote liquidity too. Open interest and trading volumes are increasing earlier. And the trading volumes reflect seasonal supply. NZX WMP futures are also providing better guidance to GDT auction outcomes.
• The success of WMP as the product vehicle is no accident in our view. WMP is the product most likely to succeed. WMP neatly sidesteps the perishability issues faced by other dairy products. WMP involves the least commitment to particular product for producers. WMP, reflecting its homogeneity, has the widest range of users too.
• NZX WMP futures meet most of the criteria for the design of a successful futures contract. The one area that needs further development is liquidity. The futures still lack the ease of transacting that would justify a “liquid” tag. In our view the industry needs to focus its attention on facilitating the NZX’s WMP futures’ journey to becoming liquid. Changes may be necessary. One suggestion we make here is to reduce the number of contracts offered within a calendar year to better aggregate liquidity. We are sure there are plenty of other suggestions floating around the industry.
• A well-functioning WMP futures market – a representative price established in a liquid market – would eventually benefit other dairy products. Such a benchmark price could become a utility for pricing all dairy products. WMP supply and demand quickly reflects any excess or shortage in milk supply so that translates into WMP prices almost immediately. With a well-functioning market, price spreads between WMP and other products would become more standardised and trade within established ranges most of the time. In time, as the reliability of the process is demonstrated, WMP futures could become the basis for fixed or floating price contracts with agreed spreads for different products.
• A well-functioning WMP futures market would have several benefits for the industry. Processors would have the means to fix the prices for a greater proportion of their throughput ahead of time. Moreover they can set the price of an easily saleable product. Consumers (i.e. food processors that use WMP in their products) too gain access to a forward price-setting mechanism. Both would also gain from being able to commit to forward quantities but without setting a price (or vice versa). Prices can be set later at benchmark prices or either party can set the price ahead of time without needing the others agreement to do so. A transaction’s price and quantity components become separated with all the flexibility that provides to both sellers and buyers.
7 December 2015
2
NEW SOLUTIONS RAISE NEW PROBLEMS
NZ has upsized the global dairy industry
The dairy expansion is the dining boom in action
NZX WMP futures have clear lead in becoming the benchmark for dairy pricing
Maturity is a liquid pricing benchmark for dairy
The New Zealand dairy industry’s flourishing into a major force on the world stage in the past decade has been source of vitality for the global dairy market. NZ’s milk production grew by about half over the decade to 2014. Yet that understates the impact of NZ’s expansion. The whole milk powder (WMP) market gives us a better idea of the scale. NZ’s WMP production expanded more than twofold over the decade to 2014. Global trade in WMP expanded by about the same amount. NZ accounts for two-thirds of that trade. Additionally, China’s share of global WMP imports was just fourteen percent in 2004. China’s share in 2014 was closer to two-thirds.
010002000300040005000600070008000
2004 2014 2004 2014
Total Total excl. NZ
Perishables WMPSMP Fluid milk
GLOBAL DAIRY EXPORT VOLUMES (kt)Source: USDA
While these are remarkable numbers their real world context is important. The WMP flow has been a success story of the dining boom. A real, and large, example of a richer world buying more, and better, food. Moreover, NZ’s expansion has moved the global dairy industry from largely servicing domestic (and often heavily regulated) markets to one that has a greater focus on cross-border trade. The dairy industry’s highly, but coincidentally, correlated geography between producer and consumer is breaking down. WMP has grown out of what is now a more distant, time-consuming relationship between the dairy farm and the ultimate consumer. The dairy industry’s big evolution is having plenty of consequences that require new solutions. One has been greater price volatility. Another has been the consequent desire of many to be able hedge against that volatility. We explore here how far we are on the journey to find a vehicle for that hedging and, in particular, how NZX WMP futures have evolved as the main candidate to be that vehicle.
NZX WMP futures’ liquidity has improved steadily since their inception in 2010. Trading volumes and open interest though remain modest by most standards so the market is easily dismissed as too illiquid. While we would not disagree with that assessment as this point in time we are choosing here to take a longer look at the market. Any futures contract will make a journey from its first listing to a destination, which, if successful, means it becomes a robust, widely-used hedging tool for a market. WMP futures are not yet at that destination but we are looking here for signs that they are on that path. Our conclusion is that the WMP futures are indeed on the path to becoming a useful pricing tool. Nevertheless the industry needs to have an active discussion about how to facilitate the next steps. Nurturing liquidity can seem like a frustrating chicken-and-egg issue. Potential participants, who are clearly symbiotic, hold back when the first-mover seemingly gains no obvious advantage. We think a “great leap forward” is unlikely. Instead we expect the market to take baby-steps, slowly flushing out pockets of participation that will eventually accelerate the WMP futures into a more liquid marketplace. We suggest some ways in which that process might be nudged along.
What do we mean by “maturity”? Or, what is a desirable outcome? In a nutshell, a liquid pricing benchmark for dairy products in NZ and Australia. Liquid would mean that participants can transact meaningfully-sized parcels in short times frames (i.e. days) without pushing prices too far against them. Benchmark would mean the pricing instrument has a relevance to a broad range of dairy participants and products.
7 December 2015
3
The discussion that follows will cover many topics. In doing so we have also, one, provided some contextual information for readers who are less au fait with the lait and, two, some digressions that are useful background but not essential to the discussion. Being mindful of that most people are busy we have greyed those segments (like this one) so they can be easily skipped.
THE AUSTRALIAN AND NZ DAIRY INDUSTRIES
New Zealand
NZ, producing just 3% of the global milk supply, is the world’s ninth largest dairy producer. However NZ takes the top spot in the dairy exporter’s league table. NZ’s dairy industry has grown at a rapid pace over the past decade. The last two years have seen consecutive records for NZ milk production. Production is expected to fall over the 2015/16 season in response to weak dairy prices and above-average levels of cow culling.
NZ exports around 95% of their milk production in one form or another. The value of NZ’s dairy export revenue (excluding earnings from cull cows) in 2013-14 was reportedly NZ$16.7 billion (A$15.2 billion) – or 33% of the total value all of merchandise exports.
2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15
Milk (million litres) 16,044 16,483 17,339 19,129 18,883 20,657 21,253
change on previous year 9% 3% 5% 10% -1% 9% 3%
Milksolids (tonnes) 1,393,000 1,438,000 1,513,000 1,685,000 1,658,000 1,825,000 1,890,000
change on previous year 10% 3% 5% 11% -2% 10% 8%
Source: Dairy NZNZ dairy production
Australia
Australian dairy production has undergone a process of consolidation over the decade and a half since deregulation. A variety of forces have driven the consolidation. Some are seasonal: periodic droughts are the main culprit here. Structural factors though have been more important. Deregulation itself is obviously one. Generic pressures, familiar to most industries, see costs persistently rising by more than prices. Industry participants all along the supply chain have thus been faced with the need become more efficient just to maintain profitability. For a few that has even meant leaving the industry altogether. The Australian dairy industry, consequently, has seen only modest production growth over the past decade or so. Productivity growth, however, has been very strong. Australia’s dairies now get significantly more milk from every dairy cow than dairies in NZ.
Australia only produces a very small share of global milk production but, like NZ, is a significant exporter. Australian dairy exports as a percentage of domestic production fluctuates somewhere within the 40-60% range. Recent Free Trade Agreements with Korea, China and Japan have seen the industry place an increasing focus on its productive capacity for infant formula and liquid UHT milk in order to better cater to these key Asian export markets.
The gross farmgate value of Australian milk production in 2013-14 was A$4.7 billion (NZ$5.2 billion). Additionally, Dairy Australia estimates that the combined on-farm and off-farm value of the Australian dairy industry to be approximately A$13 billion (NZ$14.3 billion).
7 December 2015
4
2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 (p)
Milk (million litres) 9,084 9,180 9,574 9,317 9,372 9,731change on previous year -3% 1% 4% -3% 1% 4%
Butter & AMF (tonnes) 128,379 122,486 100551 118,228 116,122 118,672change on previous year ~ -5% -18% 18% -2% 2%
SMP (tonnes) 190,233 222,484 230,286 224,061 210,964 233,835change on previous year ~ 17% 4% -3% -6% 11%
WMP (tonnes) 126,024 151,269 140,424 108,838 126,322 99,025change on previous year ~ 20% -7% -22% 16% -22%
Cheese (tonnes) 349,639 338,657 346,530 338,312 311,458 343,956change on previous year 2% -3% 2% -2% -8% 10%
*p = preliminary
Source: Dairy AustraliaAustralian dairy production
Very different product mixes
Cheese is Australia’s most important dairy product. Cheese accounted for just under one third of Australian milk production in 2014/15. Drinking milk and skim milk powder (SMP) took second and third places.
Contrastingly, NZ dominates the whole milk powder (WMP) export market and so its product mix is geared more towards producing WMP and butter. High Chinese inventories though has seen WMP demand fall away recently. As such, NZ’s production is likely to place an increasing focus on diversifying up the WMP value-chain to capture premiums for products like infant formula.
5
7
9
11
13
15
17
19
21
23
05/06 08/09 11/12 14/15
Source: Dairy Australia, Dairy NZ
AustraliaNZ
MILK PRODUCTION (billion litres)
2,500
3,000
3,500
4,000
4,500
5,000
5,500
6,000
6,500
00/01 04/05 08/09 12/13
Source: Dairy Australia, Dairy NZ
ANNUAL AV. PRODUCTION PER COW (litres)
AustraliaNZ
7 December 2015
5
WMP is in the lead for a reason
WMP is more likely to attract a “quorum” of buyers and sellers
WMP has a cash market
WMP steps around the perishability problem
WMP is modestly differentiated
WMP is the a better vehicle for processors
WMP pricing can become a benchmark for other dairy products
WMP is more likely to have a path to liquidity
WHY WMP?
Our discussion focuses on the WMP market. There is obviously a range of dairy products that could be the focus of derivatives but the most successful to date has been the NZX’s WMP futures contract. WMP’s leadership in futures volumes is an important part of the reason we have chosen it as our focus. The relative success though of WMP is not an accident. WMP looks like the product most likely to succeed.
A market must be able to gather a substantial number of potential buyers and sellers of a futures contract to have any hope of success. Without that “quorum” of buyers and sellers a two-sided market will not develop. Several factors influence whether a market achieves that quorum. Important among those are storage life, homogeneity and an activity levels in the physical market for the commodity. We think that WMP is more likely to be able to draw a quorum based on these factors. And, while there is a desire across the dairy industry and its customers to hedge a range of prices for milk (or milk solids) through to processed products (butter, cheese, milk powders etc.) many of these are less likely to gather that quorum. Parallel development across a number of dairy products is unlikely.
Milk (and many processed dairy products) are obviously perishable so their shelf-life presents a special barrier to trading them. Few businesses have the capability to store and transport perishable dairy produce. And the few that have the capability incur substantial costs to do so. That requirement naturally limits participation in those physical markets. Certainly NZX futures are cash-settled, rather than physically delivered, which limits the risk to participants. Nonetheless, for a futures market to be successful there must be a well-functioning physical (or “cash”) market for the product. Perishability precludes that from happening in many cases.
WMP is a means to step past the issue of perishability. Sugar is an example that helps explain why. Sugar is, nowadays, derived mostly from sugarcane. Sugarcane is highly perishable. Cane’s sugar content drops by the hour after being cut and there is no way to arrest that deterioration. Nonetheless, a successful futures market exists because it is not based on cane but rather the processed product: raw sugar. Raw sugar can be stored for several years so this steps past the issue of perishability. WMP obviously does not have the same storable life as sugar but it is certainly much longer than milk or many other dairy products.
Differentiation also militates against the development of a cash market in many dairy products. We discuss this in more detail later but note here that WMP has only a relatively modest degree of differentiation. The consequence is that more participants might trade the product so a quorum is more likely to form.
WMP also has the formed the greatest liquidity to date. WMP’s better liquidity though is not an accident. Firstly, WMP is the natural vehicle for milk processors to store milk as it incurs less cost and involves fewer commitments to particular products. Secondly, WMP is an obvious way to transfer dairy across borders and across time.
A well-functioning WMP market – a representative price established in a liquid market – would eventually benefit other dairy products. Such a benchmark price could become a utility for pricing all dairy products. WMP supply and demand quickly reflects any excess or shortage in milk supply so that translates into WMP prices almost immediately. With a well-functioning market, price spreads between WMP and other products would become more standardised and trade within established ranges most of the time. In time, as the reliability of the process is demonstrated, WMP futures could become the basis for fixed or floating price contracts with agreed spreads for particular products.
The sequence of events to get to this “single benchmark model” would take many years to develop. And that model certainly looks and sounds messier than the multiple benchmarks model that would result from parallel market development. To be clear, we too would like there to be liquid benchmarks in many processed products. Our point is that we cannot see a path to that multi-benchmark world but we can see a potential path to a single-benchmark world. One final qualification we would make is that our thinking so far only applies to the processed dairy products. The fluid milk market might well be a different matter given that it introduces many potential new players (dairy farmers) to the market and opens up the processors’ buy side too.
7 December 2015
6
What is the state of play in NZX WMP futures?
There are influences aplenty
WMP futures trading volume are growing
Trade-by-appointment no longer
A CLOSER LOOK AT THE NUMBERS
We take a closer look at the trading data for NZX WMP futures in the following sections for signs that the market is on the path to maturity. We also note the underlying seasonality of dairy production. And we briefly summarise the development path of a recently matured futures market (for European Milling Wheat) to provide a point of comparison.
The data will be affected by a multitude of factors. Some of these will be the supply-demand balance in the market at the time. Different price levels are also likely to affect participation. And, of course, there will be circumstances peculiar to any a particular time that have an influence too. Some of these are interesting of course but would only be digressions here. We are not attempting to provide an historical account of the WMP futures market. Our narrower aim is to investigate two principal questions. One, is the WMP futures market is on a path to “maturity”? And two, is the WMP futures getting closer to a time when it is robust enough to serve as tradeable benchmark for a broader range of risk management tools? As such we will look past the data’s “wiggles” through to the broader trends that affect these issues.
Volumes
WMP futures trading volumes has grown substantially (see next chart). 2015’s (year-to-date) trading volume, at around 96,000 contracts, is seventeen times the volume of trades in 2011. The turnover is now equivalent to about seven percent of NZ WMP production.
0
2,000
4,000
6,000
8,000
10,000
12,000
14,000
16,000
18,000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2011
2012
2013
2014
2015
WMP FUTURES TRADING VOLUMES*
Source: Bloomberg
* # contracts
Contract Month
Trading is gaining in regularity too. NZX WMP futures only traded intermittently when they were first listed. That intermittent trading pattern continued for quite some but has changed in the past eighteen months or so. Trades will typically occur on 3-4 days of any 5 day windows (see next chart). And in busier periods there are trades on most days. While there are still days when there is no trading in particular contracts those (non-) events are now themselves tending to become intermittent. Ideally there would be trades every day so, despite substantial improvement, the market still has a little way to go on this measure.
7 December 2015
7
Dairy’s seasonal pattern is embedded in WMP futures
0%
10%
20%
30%
40%
50%
60%
70%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
20112012201320142015
3 Months to Expiry: % Days With Trades
Contract Month
Source: CBA
NZX lists a futures contract in each calendar month for several different dairy products. Uniformly improving liquidity across contracts of different dates would be useful – but it is unlikely. Instead we should expect the seasonally-important contracts to develop liquidity first. An agri-commodity futures market, as one phase of development, should see a seasonal pattern in trading volumes. The seasonality will reflect the pattern of supply over the year. In crops that is at harvest time. For dairy it will be while cows are in milk. Dairy seasonality is discussed separately (see next box) but here we will simply note that futures volume seasonality has indeed emerged (see next chart). NZX WMP futures trading volumes have evolved such that about seventy percent of the trading is in the September-February window when NZ milk supply is heaviest.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
2011 2012 2013 2014 2015
On: Sep-Feb
Off: Mar-Aug
TRADING VOLUMES: ON- & OFF-SEASON*
* % annual trading volume
Source: CBA
7 December 2015
8
WMP futures open interest is also growing
Seasonality
Milk production is highly seasonal in any given dairy region of the world. NZ’s milk production exhibits a strong seasonal pattern. Production peaks in the period from September to January. This seasonal pattern is reflected in futures contract participation across the year. NZX WMP contracts in that window see much more trading – as we should expect. The demand for hedging in many agri-commodities initially comes primarily from the supply side. Consequently we should expect more participation when supply pressure is greatest – i.e. when intermediate holders are under the most pressure to clear current inventories before new inventories arrive.
0
50,000
100,000
150,000
200,000
250,000
300,000
Jan Feb Mar Apr May May Jun Jul Aug Sep Oct Nov Dec
2011
2012
2013
2014
2015
NZ MILK PRODUCTION*
Source: DCANZ
* kgs of milk solids
Open Interest
WMP futures open interest has also grown substantially since the contracts were first listed in 2010. Increases of three and four times are common among the seasonal peak contracts. The odd one of those has gained much more. Growth has particularly accelerated in the past two years (see next chart).
0
1000
2000
3000
4000
5000
6000
7000
8000
JAN FEB MARAPRMAY JUN JUL AUG SEP OCTNOVDEC
2011
2012
2013
2014
2015
WMP FUTURES - MAXIMUM OPEN INTEREST
Contract Month
Source: Bloomberg
7 December 2015
9
Trading now starts earlier
The growth has been accompanied by another encouraging pattern. Open interest, and so trading, is rising earlier (i.e. with more time to expiry) for most contracts (see next chart). A greater volume of earlier trading will allow hedgers to manage their risk for a longer time period and so make the market more useful for them. That can only encourage more participation. Finally, the pattern of open interest also reflects the seasonality we discussed earlier.
0
1000
2000
3000
4000
5000
6000
022446688110132154176198220242
2011
2012
2013
2014
2015
WMP FUTURES - OPEN INTERESTOctober Contract
Days from expiry
Source: Bloomberg
European milling wheat – an example for comparison
European Milling Wheat (EMW) Futures are listed by Euronext to facilitate trade in milling grade wheat of EU origin. The EMW or Paris contracts today have open interest that rivals the Chicago wheat futures contract (i.e. the most liquid contract). The underlying wheat quantities are substantially smaller but the Chicago contract does have receive a boost to volumes from Trading volumes¸ at 15-20,000 per day, are more modest but can easily accommodate substantial transactions. EMW futures though have grown from humble origins. The contracts were first listed for trading in 1999. The initial years saw very modest volumes traded and open interest never got larger than several thousand lots. Trading was intermittent too. Trades were occurring on only three or four days per week. Open interest and trading volumes nonetheless continued to climb steadily, though not spectacularly, for several years. 2007 – some eight years after launch – was the year when participation accelerated sharply and EMW futures started on the path to become a much more liquid market. Trading volumes are now equivalent to about five percent of EU wheat production. The distribution of EMW futures liquidity is also worth highlighting. The November future was the harvest contract. Consequently the November contract saw the earliest rise in volumes and that has grown to be the most heavily-traded. Indeed EMW futures have now been so successful that the November contract has been replaced by two contracts (September and December expiries). More importantly though, the example shows that a new futures contract can be successful but it can take quite some time to build liquidity. While we do not expect NZX WMP futures to reach a similar level of liquidity any time soon the example demonstrates that success often takes time to accumulate.
7 December 2015
10
WMP futures are providing a better guide to GDT auctions
0
50,000
100,000
150,000
200,000
250,000
1999 2001 2003 2005 2007 2009 2011 2013
JAN
MAR
MAY
NOV
EU MILL WHEAT - MAX. OPEN INTEREST
Source: Bloomberg
0
5
10
15
20
25
30
35
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Chicago
Paris
OPEN INTEREST: PARIS v CHICAGO*
* December 2015 contracts from start 2015
Source: Bloomberg
Auction convergence
The NZX futures are available to trade each day while GDT auctions only occur bi-monthly. Traders will of course make physical trades between GDT auctions but these are much less visible. The WMP futures can thus provide a more transparent indication of where the market is trending between auctions. Initially the WMP futures were playing that role only poorly, if at all. The WMP futures tended to meander between auctions and then abruptly re-set to levels set at the GDT auctions. WMP futures are now performing that role better and have been improving since late 2013 (see next chart). The precise measure is the (absolute) difference between the last traded price in futures before the second monthly GDT auction and the subsequent settlement price. The divergences in the period 2011-13 were mostly in the $100-$200 range (or about five percent). The same numbers in 2014 and 2015 are around $50 (or about two percent). Clearly we can see that futures trading leading up to settlement is now taking futures prices closer to where the physical market is trading. Thus the futures market is matching the GDT auctions when it matters.
7 December 2015
11
WMP futures performed well during recent stress points
-1500
-1000
-500
0
500
1000
1500
Nov 10 Nov 11 Nov 12 Nov 13 Nov 14 Nov 15
Source: CBA
WMP FUTURES AUCTION CONVERGENCE
* NZX Futures contract price
difference between pre-GDT auction
close and post-auction
settlement
Very recent auctions have seen some larger divergences. That has coincided with considerable volatility in the underlying dairy markets (i.e. they were not solely dislocations in the futures markets). That can be looked upon in two ways. One is that the market is still not quite mature enough to easily deal with the stresses that were evident in dairy markets at that particular time. The other is that the WMP futures dealt with these stresses with, by past standards, small final divergences. So what is remarkable is not that there were significant divergences but that those divergences were not much larger. The performance is actually a sign of progress. So, while we do not view the WMP futures market as being mature, we do see these recent episodes as a sign the market is on the journey to maturity.
7 December 2015
12
Are WMP futures structured to develop to maturity?
Seven success factors
WMP futures are better than any current alternative
Liquidity though is the big issue
DESIGNED FOR SUCCESS?
The data suggests that the NZX WMP futures are maturing as a market. The next question to ask is whether the market, given the design of the contracts, is likely to reach maturity. Here we outline some criteria for success and discuss whether there are any factors that might prevent that occurring. And we also suggest some steps that might help with the maturation.
A list of success factors, distilled from the academic literature, is shown in the table below. We will discuss each of these facets in the context of NZX WMP futures.
1 BestA physical market must be more closely tracked by its own futures contract than any alternative hedge.
2 CostThe liquidity cost of using the own futures market should not be materially greater than any alternative hedge.
3 Active The physical market must see frequent transactions.
4 SizeThe physical market must be large enough to support a large pool of potential hedgers and speculators.
5 VolatilePhysical prices must be variable enough to create hedging needs and speculative interests.
Structure 6The chain of production cannot be either vertically integrated or highly concentrated.
Homogenous 7The commodity must be homogeneous or at least have a well-understood and objective grading system.
Alternatives
Physical Market
FUTURES CONTRACT SUCCESS CRITERIA
Source: Brorsen, B. Wade & Fofana, N'Zue F., 2001. "Success And Failure Of Agricultural Futures Contracts," Journal of Agribusiness, Agricultural Economics Association of Georgia, vol. 19(2)
Better than nothing?
The first couple of criteria ask the seemingly obvious question of whether a futures contract is better than the best alternative hedge. The alternatives, if they even rate that description, are underwhelming. Skim milk powder (SMP) futures are available in the US and Europe but neither is likely to be a more effective hedge. Firstly, the contracts are for skim, not whole, milk powder. Secondly, the locations are obviously different. Both differences would very likely mean the basis (or spread) to WMP in New Zealand or Australia would be highly variable and so perform ineffectively as a hedge. The liquidity of both contracts is inferior too. So, yes, WMP futures dominate the alternative hedges. The lack of an alternative though should not be a surprise. As we mentioned in the introduction, this market is still very new so some of the foundations we take for granted in other, older markets are still being laid.
Liquidity costs (the second criterion), by contrast, do require some further discussion in our view. While there is no realistic alternative to NZX WMP futures – so they are the lowest cost alternative – that does not mean liquidity costs are low. WMP futures have seen big price swings in the past year or so. We think that much of those swings reflect the waxing and (mostly) waning of the market’s perceptions about fundamental conditions – but not all of it. We suspect a proportion is simply that the market only accepts very modest transaction sizes before prices have to rise or fall substantially to find the next seller or buyer. What that means is that, for a material-sized transaction, the gap between the first price and final price when hedging in the WMP futures is large compared to other agri-commodity markets, i.e. the liquidity cost is sizeable. Here is
7 December 2015
13
WMP has an active cash market…
…that has a high and growing value…
…with volatile prices!
The GDT auction system is a bonus at this stage of development
Does the WMP have a suitable industry structure? The industry is not too vertically integrated
where the “better than nothing” question is real. Few players will want to burn five to ten percent of the value of a transaction just to set a hedge. The liquidity cost needs to be reduced further over time to encourage more participation. We make one suggestion form improving liquidity in a later section (Too much too young).
Derivatives must be derived from something
The physical market is crucial in the success or failure of a futures contract. The third to fifth criteria focus on the physical market.
Unless there are a substantial number of transactions in the physical (or “cash”) market then there is little need to hedge them. Thinking about the alternative can make that concept clearer. A market where fixed price contracts (of substantial duration) are the predominant vehicle will have many transfers of the commodity but few transactions where price is negotiable. The physical WMP market is an active market so the activity requirement (criterion three) is easily met.
Size obviously matters in this respect too. New Zealand and Australia produce about 1.5 million tonnes of WMP each year. Production now is worth, depending on price, $US3-5 billion. Volumes, and consequently values, will continue to trend higher in the years to come too. The volume of production is clearly large enough to generate many transactions. And the value of the transactions is also high enough to mean the transaction costs in liquid markets would be minimal (if not trivial). The physical WMP market clearly meets criterion four. Note too that these numbers only include the WMP market itself. The potential is greater if the market develops such that WMP becomes benchmark price for the broader suite of processed dairy products.
The volatility requirement (criterion five) is, as anyone who has followed dairy market in the recent years will know intuitively, is easily met! Global Dairy Trade (GDT) auction-to-auction volatility since their inception in 2008 has been around 5-6%. That auction-to-auction volatility is around 30% in annualised terms. Volatility at that scale in physical WMP prices is a match for any agri-commodity and certainly for those with successful futures markets. In short, there is plenty of volatility against which participants might want to protect themselves.
The NZX WMP market has the added bonus of the GDT auction pricing system to tie the physical and futures markets together. The GDT auction system offers three advantages. The first is that there is a direct link between the two markets because the auction price sets the ultimate futures settlement price. So the futures are anchored to a physical price despite not being deliverable. The GDT auctions though offer more general transparency benefits. The GDT auctions are an excellent mechanism to generate an explicit cash price available to all. Moreover, prices are not just for spot transactions, but also for several months forward. The GDT auctions also offer the further valuable bonus of being based on actual and verifiable transactions. Most agri-commodity markets do have physical indicator prices that are either known to the trade or published via various price reporting services. And, while they may be more frequent, few of these will match the GDT auctions for transparency.
In our view, the physical WMP market presents no barriers to the futures market developing.
Is there an elephant in the room?
The fourth criterion concerns the structure of the industry and whether that is might prevent a futures market developing.
The first issue within this category concerns vertical integration. Vertical integration refers to the extent that different steps in a production process are done within one company. If the production of WMP was undertaken by the same company to use as an ingredient in other products then there would be no external transaction. If this
7 December 2015
14
Is Fonterra’s NZ dominance an issue?
Fonterra’s has a greater interest in a well-functioning market
Fonterra’s “control” is heavily qualified
Overall uncertainty about supply is also a factor
pattern was widespread across the industry then there would a high degree of vertical integration (Case B in diagram on right). Without external transactions there is obviously going to be no need for hedging. The dairy industry is characterised by some degree of vertical integration but, for now anyway, that is modest as far as WMP is concerned. Vertical integration is unlikely to be a barrier to the development of a futures market.
Concentration, by contrast, is perhaps more of an issue. Whether that is a real or perceived issue is a matter of conjecture. Fonterra is the elephant in room. Fonterra accounts for large proportion of the physical trading in NZ dairy products. And they clearly have an impact on GDT auctions through their decisions about the volumes and the product mix they will offer. At Fonterra’s size that influence is unavoidable. And their influence would be substantial regardless of whether or not the GDT and futures pricing mechanisms existed.
o
Many are at least wary though that Fonterra might use that influence malignly. This wariness can be a barrier to market growth if the lack of trust stops potential users – hedgers or speculators – committing to the market.
People are, quite rightly, reluctant to rely on the benevolence of others to regulate their behaviour. Conspiracy theories breed easily in this context and they are near-on impossible to confirm or refute. The argument against these theories though is not really based on Fonterra’s goodwill but rather its self-interest. Fonterra is likely to benefit far more from a well-developed derivatives market than it could from manipulation of either the GDT auctions or NZX futures market for a number of reasons. Firstly, manipulation would not be a sustainable activity. Either (or both) of the GDT or NZX markets would quickly wither as any manipulation would quickly become evident. And, with that, the number of opportunities to make money from the manipulation is likely to be small. Secondly, the size of either the auction or futures market would limit the position size (and so potential profits). The “prize” from manipulation is small so the question arises: if manipulation is the goal, why would Fonterra bother with formal markets at all? Surely it must be easier, cheaper and safer to do that in more opaque, bilateral markets.
Fonterra’s level of “control” over the market seems to be taken for granted but can easily be overstated. Without control, manipulation is much easier said than done. Fonterra is a milk aggregator and processor so it makes money primarily on throughput. The salient point is that Fonterra makes money not by owning milk but by moving milk (and milk products). Fonterra, with such a big market share in New Zealand, can make substantial profits from this business. Fonterra’s control over their volumes over any short period of time though is limited as they largely have to accept the amount of milk that contracted dairies send to them. That lack of control over input supply – which must be processed because it is perishable – means Fonterra’s control over their output is also limited in the short-term. That is true whether supply is easy or tight. Fonterra cannot conjure more milk if farmers cannot supply it (because, for instance, there is a drought in NZ). Fonterra will have limited control over market if they have limited control over their own output.
Another aspect of this lack of control is also important. Fonterra’s lack of control is driven by a high degree of uncertainty about supply. The high degree of supply uncertainty is common to agricultural markets and is primarily driven by the weather. Weather is in the lap of the gods and so greatly levels the information playing field between those who are in the supply chain (hedgers) and those who are not (speculators). This more level
7 December 2015
15
Market dominance is unlikely to be an issue but first impressions die hard
WMP needs to be “homogenous enough”
Many dairy products are “not homogenous enough”
WMP’s differences are not salient
Is the WMP market’s monthly contract schedule holding back liquidity?
playing field is what encourages speculators that help provide liquidity to the market.
In summary, Fonterra’s desire and ability to manipulate the market is highly questionable. Nevertheless, only time, and consistent behaviour through time, will practically eradicate qualms about Fonterra’s role. Fonterra’s place in the NZ dairy marketplace though should not ultimately be a barrier to broader participation in the WMP futures market.
Same, same but same enough
Homogeneity is the final criterion. A futures contract can only be successful if the product is homogenous. We asserted earlier that WMP is homogenous. We outline the reasons for that assertion here. All goods and services can be placed somewhere in the heterogeneity-homogeneity spectrum. Consequently, homogeneity is a matter of degree rather a simple categorical quality. So a product, like WMP, simply needs to be “homogenous enough” not “absolutely homogenous”.
Many dairy products are not highly homogenous and it can be helpful in understand what is meant by “homogenous enough” by describing what is “not homogenous enough”. Cheese is an obvious example with its great many forms and flavours. Even for a common cheese variety, like cheddar, there is a spectrum running from low-cost “rat trap” to the more Epicurean varieties. The consumer will not see one of these products as a close substitute for the other. Consumers can be finicky of course but even where the vagaries of their choice processes are removed, and only the end result matters, the differences remain. Chefs, for example, effectively perform this role when selecting ingredients for their dishes. A pizza chef will not keep customers for long if she makes pizzas with cream cheese. Nor will a pastry chef if he makes cheesecake with mozzarella. A chef can substitute between different brands of the same cheese but their ability to substitute across varieties is limited. The implication is that the price for any particular variety of cheese has limited relevance to another variety. In a sense this brings us back to the size of the market. Mozzarella cheese is no doubt a lucrative market but it nonetheless a specialised product. Consequently the number of players, and the volumes and values, is relatively small. Same, same but different!
WMP differs in most of these dimensions. We have already described the scale of the market in volume and value terms. The number of participants is also much larger because WMP is a product with a wide variety of uses. Some will object that WMP does have a degree of differentiation. That is true but the issue here is the degree of differentiation (or heterogeneity) not differentiation per se. The product in question simply needs to be far enough down the homogenous end of the homogenous-heterogeneous spectrum. “Commodity” differentiation is greater than many realise. Crude oil, a market with a highly successful futures market, is an example. Crude oil has several differentiating qualities. One is sulphur content – the difference between “sweet” (low) and “sour” (high) crudes. Both the sweet-sour spectrum can be refined into various products (gasoline, fuel oil etc.) such that end-consumer is indifferent to the grade of crude oil. Similarly, different types of WMP can be used by food manufacturers to produce, via the technology of their processes, products with no salient difference for the final consumer. This degree of differentiation doesn’t matter so the users can shift between varieties to a greater extent. That means the game covers more players with greater volumes and values. Same, same but same enough!
Too much too young?
We have already mentioned that NZX lists a contract for each calendar month for several dairy products. Here we discuss the pros and cons of that futures contract frequency. Is dairy a market that is better suited to such a high frequency (monthly) futures expiry? Or, does frequent expiry dissipate liquidity?
Our discussions with various participants raised several suggestions as to why the expiry pattern should be frequent for dairy produce.
7 December 2015
16
A sustained large flow of a commodity doesn’t seem to predict how many contract expiries there will be in year
Perishability is not an issue for WMP
Physical and futures do not need to align and in practice they rarely do
One reason suggested was that milk production has a long season, and the volumes are at a high level for much of that time, so the market needs a more frequent hedging mechanism. Consequently less frequent expiry dates would not be enough to capture the flow of production as it comes to market. Other agri-commodity markets have large flows over several months but do not have monthly futures expiries. Sugar is an example. Southern hemisphere mills produce the bulk of their raw sugar in the July to November period. Most of that sugar is priced against two futures contracts: one that expires at the end of September and another at the end of February. Palm oil provides a contrasting comparison. Palm oil is produced year-round but, like milk, it has large seasonal swings in output. The associated futures contracts eventually have an expiry for each month of the calendar year. Two such contrasting examples suggest that a sustained flow of production probably does not do much to determine the frequency of futures contracts.
Dairy’s perishability is also proffered as another reason for frequent expiry. We are somewhat sceptical of this argument. Firstly, the NZX WMP futures are cash-settled, not deliverable, so the perishability or otherwise is not directly an issue (even though it is obviously relevant in the physical market). Secondly, perishability varies materially across dairy products – milk, butter, cheese and milk powders all have substantially different storage lives. Consequently, we can perhaps see why milk and other products with similar lives might need more frequent matches with futures contracts. For products with longer lives (like WMP) perishability is not really a constraint for the range of frequencies we are discussing here. In our view perishability is perhaps more of an argument for the right frequency rather than high frequency per se.
Finally, the GDT running auctions twice a month, or perhaps them originally being run monthly, was another suggested reason. NZX WMP futures monthly contracts provide a closely-linked hedge to prices in that month. Yet we are sceptical that frequent physical transactions means the associated futures contracts must also have matching or frequent expiries. Many of the most liquid agri-commodity futures contracts do not trade any more than a half-dozen contracts in a year (see next table). The CBOT corn contract, probably the most liquid among this group, only lists five contracts per year. The soy complex has something more akin to the processing stage in dairy products. Soybeans and its products, soymeal and soy oil, trade seven, eight and eight respective futures expiries per calendar year. Looking at agri-commodity futures more broadly, very few futures that are transacted as hedges against physical positions are targeted at a particular expiry date. Traders enter a great many physical trades do not have a definite exit date. Traders also frequently transact hedges that they know will very likely be unwound well before the futures expire because the corresponding physical positions will be closed well before that time. The main lesson here is that futures contracts do not need their expiries to match closely with physical pricing dates to be a useful hedge.
FUTURES CONTRACT FREQUENCIESJan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec #
WMP NZX 12
Corn CBOT 5
Wheat CBOT 5
Canola ICE 5
Soybeans CBOT 7
Soymeal CBOT 8
Soy Oil CBOT 8
Sugar ICE 4
Cotton ICE 5
Live Cattle CME 6
Feeder Cattle CME 8
Source: Bloomberg
7 December 2015
17
Fewer WMP expiries could combine pots of liquidity
And, the whole might sum to more than the parts
Fewer expiries would require a change to their settlement pricing mechanism.
WMP futures are the right vehicle…
…but it needs to be fuelled by more liquidity
Further to the last point though we would like to raise the possibility that having fewer WMP expiries in a year might actually improve liquidity. Our contention is that, by having fewer expiries, transactions would be concentrated in fewer contracts and thereby increase liquidity (and reduce liquidity costs). Moreover it is quite possible, even likely, that the liquidity would soon improve by more than the just the sum of the trades of two or three contracts shrinking to one.
There are several reasons why the whole might prove to be more than the sum of the parts. If traders are generally not targeting particular dates for unwinding trades (and so are indifferent between a few different futures expiries), then extra contracts will dissipate liquidity because it decreases the chances that two traders will be looking for counterparties in the same contract. Moreover, even when particular dates are selected, the hedge might be just as effective in a futures contract that has a slightly earlier or later expiry. Consequently the hedges will still be transacted but are likely to find counterparties sooner. Speculators are a much maligned group but they are an important source of liquidity in trading markets. Some speculators will have their own fundamental views about WMP prices for months ahead and position themselves accordingly. Others will have much shorter time horizons. One speculator might have a view that prices are going to fall until next Tuesday when she will buy her short back. Another speculator might be keen to sell next Tuesday because his trading strategy dictates that he sells the week before a GDT auction. These two speculators are more likely to find one another if they are trading the same futures contract. The other more general point to make is that as the number of different strategies increases the likelihood of being able to transact also increases. This would nudge liquidity into a better place and probably make more transactions worth attempting.
Moving to fewer contracts though would probably require a change to the settlement pricing mechanism. Currently the settlement price for any contract is the average of the two GDT auctions in that month. If contract expiries were spaced even two months apart an average of prices over those two months might prove problematic. After the first few auctions the final auction would have only a 25% weight in the final price. The contract would then have little efficacy as a hedge and would more than likely become a lame duck. Using a single GDT auction as the settlement mechanism would avoid this issue.
JUST ADD LIQUIDITY
The dairy industry and their customers have experienced substantial price volatility in recent years. That volatility has generated a desire for risk management tools. To meet that desire though a liquid, tradeable price benchmark must emerge.
Our view is that NZX’s WMP futures are the most likely vehicle to first provide that liquid benchmark price. As we have discussed here, WMP futures have already taken substantial steps down that road. And they have a clear lead to any other alternative. WMP futures take the lead in the benchmark race for good reasons though, rather than just by accident. The future new consumers of milk (direct and indirect) do not live next door to the dairies that supply the milk, as most past consumers did. WMP is a vehicle that steps around the perishability problem that is more pressing as cow and consumer are separated by greater time and distance. Moreover, for milk processors and users alike, WMP offers the least need to commit to particular products and so is the natural product for these parties to exchange. While we’d agree that having benchmark prices for a range of dairy products is desirable the practicalities mean it is unlikely such a suite will emerge simultaneously with liquidity.
Desire alone though will not create liquidity. The NZX WMP futures have taken substantial steps in the right direction but there is still some way to go. Steps need to be taken to improve liquidity to ensure that journey continues but that requires continued active discussion. The discussion should not though become a search for a panacea that will quickly increase liquidity. The discussion should focus on the small steps that will add to liquidity – a call to evolution not revolution. GDT provides an example with their continuous refinements to their auctions terms and conditions. Our suggestion that fewer contracts in the calendar year might promote liquidity is made in this context. Fewer contracts is one among potentially many suggestions that might facilitate greater liquidity. Each change though will invite a few new participants to the market and allows existing participants to scale up their activity. Robust liquidity can emerge in time.
7 December 2015
18
ASB Economics & Research Phone Fax Chief Economist Senior Economist Senior Economist Rural Economist Economist
Nick Tuffley Jane Turner Chris Tennent-Brown Nathan Penny Kim Mundy
[email protected] [email protected] [email protected] [email protected] [email protected]
(649) 301 5659 (649) 301 5853 (649) 301 5660 (649) 448 8778 (649) 301 5661
(649) 302 0992
https://reports.asb.co.nz/index.html @ASBMarkets
ASB Economics ASB North Wharf, 12 Jellicoe Street, Auckland
Important Disclosures
This document is published solely for informational purposes. It has been prepared without taking account of your objectives, financial situation, or needs. Before acting on the information in this document, you should consider the appropriateness and suitability of the information, having regard to your objectives, financial situation and needs, and, if necessary seek appropriate professional or financial advice.
We believe that the information in this document is correct and any opinions, conclusions or recommendations are reasonably held or made, based on the information available at the time of its compilation, but no representation or warranty, either expressed or implied, is made or provided as to accuracy, reliability or completeness of any statement made in this document. Any opinions, conclusions or recommendations set forth in this document are subject to change without notice and may differ or be contrary to the opinions, conclusions or recommendations expressed elsewhere by ASB Bank Limited. We are under no obligation to, and do not, update or keep current the information contained in this document. Neither ASB nor any person involved in the preparation of this document accepts any liability for any loss or damage arising out of the use of all or any part of this document.
Any valuations, projections and forecasts contained in this document are based on a number of assumptions and estimates and are subject to contingencies and uncertainties. Different assumptions and estimates could result in materially different results. ASB does not represent or warrant that any of these valuations, projections or forecasts, or any of the underlying assumptions or estimates, will be met.