The Implications of NZ Agriculture's Debt - Update June 2009

June Update
We earlier looked at the implications of dairy farm debt to September 2008 (further below). That work was based on a sophisticated approach estimating the present values of farms based on their profit under normal or optimised production – something a little too complicated for many to wade through. While the approach is still valid it also requires work on costs for the 2009-2010 season – something not yet completed.

This update takes a simpler approach - still using that earlier work but with updated debt figures and ignoring farm present value calculations. We provide data for a range of three market asset values and rely on the reader to select the value matching their perceptions, or to interpolate.

Since September 2008 there have been dramatic falls in dairy asset values – farms being the major component. Changes in cow prices are also significant. Fonterra share prices have dropped again this season but still appear to be overvalued at $4.52 per share given their net tangible asset value of 78 cents as of the 31st of January 2009.

No suggestion should be made that current lending to agriculture is for the development of assets. Farms on average are in all probability going backwards for all forms of maintenance.

Banks are continuing to lend to agriculture but mostly for ubiquitous organic growth. It would appear that the majority of lending is disappearing into properties already in negative equity.

While the dairy industry is clearly the major concern, significant lending is also going to other sectors of agriculture not immune to many of the same issues. Problems with excessive lending to agriculture are thus greater than suggested by dairying alone.

Broad implications flow from the above, and include:

1. The obvious – at least 20% of NZ’s dairy farm production can not, and will not ever, meet their debt servicing commitments even under the most promising payout and interest rate scenarios.

2. Reduced stability of the NZ financial system. Banks are continuing to lend taxpayer guaranteed funds to farms with no prospects of ever repaying the debt. That would appear to be high risk and reckless in terms of the financial stability of the banking system. That risk exposure is flowing through into higher bank charges and interest margins as banks struggle to retain profitability and/or viability.

(In an open economy shouldn’t we expect the emergence of new banks without these toxic assets and high offshore borrowing costs given they will be more competitive and make higher profits? That raises questions as to whether the RBNZ’s approach to financial stability is one of maintaining existing banks rather than encouraging a wider range of financial institutions, or new ones. Some consideration is due of whether the biggest threat to financial stability comes from having all the major banks working to what is essentially the same model. I suspect Basle risk models don’t consider this aspect.)

3. Increasing misallocation of capital within agriculture. Capital is short and is being rationed but not on the basis of risk or potential industry profits. It appears financial institution lending is preferentially going into propping up farm asset values (and maintaining lender’s profits on paper at least) while more productive uses are missing out. Included in the latter could be finance for new processing capacity e.g. competition to Fonterra.

4. Artificially maintaining higher farm asset values continues to delay the adjustment agriculture must make to a lower capital cost farm structure, and will distort industry strategies.

5. Given the scale of financial institutions’ problems with agricultural debt, and that it doesn’t to any extent yet show in bank disclosure statements, combined with the RBNZ’s limited admonishment of banks in regards to their agricultural lending, suspicions arise as to the true visibility of financial institution problems.

The full November 2008 Analysis:
Covered:Agricultural Debt; Situation regards Income; Ability to Service Debt; Valuation of Assets; Value of Dairy Farm Assets; Farm Solvency and; Conclusions regards NZ's Agricultural Debt.

Agricultural Debt
The RBNZ has in its last two Financial Stability Reports provided information that together allows a better estimate of dairy farm debt than has previously been possible. The May 2008 report established a pattern for the distribution of debt for sheep-beef and dairy properties from analysis of data from IR10 tax returns for the year ending March 2006. The RBNZ information illustrates a concentration of debt with 20% of farms - particularly the new, higher gearing for dairy than sheep and beef, and debt being 28% higher than reflected in the banks Standard Statistical Returns (SSR).

The November 2008 Report provided accurate information on the breakdown of dairy and sheep-beef farm debt from the RBNZ's SSR as of 30/06/08. Combined, with some guesswork as to how debt is distributed in relation to production based on data from other sources, allows the data Figures 1-3 and in Table 1 to be proposed as representing likely farm debt levels as at the end of September 2008.

MAF surveys of agricultural debt in 1988 and 1998 established that sheep and beef farms had much higher levels of family debt than dairy farms. Other providers of lending to agriculture include finance companies for stock mortgages and initial start ups for dairy conversions.

The May data from the RBNZ only had dairy debt only one third larger than sheep-beef debt, and clearly under represented. The approach taken to establish dairy farm debt as of 30/09/08 was to scale the 30/06/08 data by the change in SSR for agriculture (4.8%) and then add 10% for all other sources of finance (taking the figure to close to $27.5 billion - 61.49% of all agricultural debt).

For sheep and beef farms, the figure in the RBNZ's May report was simply scaled by the increase in SSR lending for agriculture for the 30 months to the 30/09/08. The resulting figure for sheep and beef farm debt is $14.7 billion.

Table 1 Original Data Source: Reserve Bank of New Zealand

How should these figures be interpreted? The 358 most indebted dairy farms have average debt of $17.3 million, while the 589 most indebted sheep and beef farmer's average debt is $7.9 million.
Fifty percent of dairy farm production has only 10% of total dairy farm debt while the top 2.5% alone has another 10% or $2.7 billion - 20 times that of the least indebted 50%.

Situation regards Income
It is well established that NZ agriculture has high levels of debt, but less acknowledged that 50-60% of productive farm capacity has little or no debt. Farms with little or no debt are not our primary concern here and now, although questions remain as to whether even debt free NZ farms will be viable in future. That leaves us looking at the issues facing farms with significant levels of debt:

1. Can the debt be serviced?
2. If the debt can't be serviced, is the business still solvent i.e. does the value of security cover the level of debt?
3. Are these just transitory issues, or are they systemic?

The answer to the first two questions largely depend on farm incomes - first to service debt, and second to support the value of the assets secured. A close look at the relationships between debt and income in agriculture will challenge much that has been taken as gospel in NZ for decades. We get into areas where discussion engenders conderable discomfort, there are questions surrounding the basis for industry attitudes, and the quality and independence of some information provided.

The answer to the third question requires an article in its own right.

MAF's annual Situation and Outlook for Agriculture provides the information on which most government planning is based.
Agricultural Sector Income from Situation and Outlook for NZ Agriculture and Forestry, 2008 is plotted in Figure 4:
Figure 4 Data Source: MAF Situation and Outlook

The MAF data is supported by the following commentory:

The situation within individual sectors of New Zealand agriculture can differ markedly. Gross agricultural revenue at the farm gate is estimated to have risen by 6.4 percent during the year ended 31 March 2008. This is due to the increase in dairy revenue being greater than the decrease in meat and wool revenue.

However, the amount of interest paid by the agricultural sector rose by more than 30 percent in the year ended 31 March 2008. As a result, agricultural sector income – an aggregate measure equivalent to the overall agricultural sector’s farm-gate profitability – is estimated to have gone down.

In coming years, agricultural sector income is forecast to improve. This is due to dairy gross revenue remaining strong, a recovery in farm-gate meat prices, a depreciating exchange rate and falling interest rates keeping a lid on interest payments.

There is valuable information contained in MAF's Situation and Outlook but equally some is simply not credible. This forecast suggests some semblence to what bank managers working in agriculture are likely used to seeing. Declining income to date, but a reversal forecast with improvements continuing into the future.

The following is selective data extracted from MAF Regional Dairy Models:

The purpose of this data is to illustrate details regards dairy farm incomes, but some is also relevant to the question regards what information MAF provides is credible or not.

Starting from 2008 - the best year in 40 for dairy prices, they then expect farmgate incomes to increase over coming years.
This article is centred around the major issue of agricultural debt, but MAF, and possibly also Treasury, have issues with this subject. MAF conducted major surveys of agriultural debt in 1988 and 1998. They have farm models they update annually that include liabilities and interest. They are aware of the rapid expansion of agricultural debt from figures provided by the RBNZ. Illustrated here: NZ Agricultural Debt

Using RBNZ data we established a level of dairy farm debt as of the 30th September 2008 equal to an average of $21.01 per Kg of milk solids production. MAF however forecasts regional average dairy farm debt per Kg milk solids production at the start of the 2009 season as ranging from $7.51 in Northland to $14.29 for Southland - numbers that are not even close.
The interest paid figures MAF uses in its 2008 Situation and Outlook appear to be about $1 billion short. Given that the agricultural sector income figures from 2007 to 2010 are between $0.844 billion and $1.033 billion this would make agricultural sector income negative over the period.

An extract from MAF's 1998 report on Agricultural Farm debt may be informative:

One very large dairy farm returned a form showing total liabilities more than 3 times greater than any other farm, and was removed from the sample as an outlier.

Alternative analysis might suggest that this was a fortunate form identifying key data rather than being an outlier, that no farm is likely to have three times the debt of the next most indebted, and that information on highly indebted farms was likely poorly represented in the data sample.

The MAF model behind Table 2 works on a year to 31st March. This complicates analysis when most farms and Fonterra work to balance dates of May, June or July.

The data in Table 2 has been selected and rearranged to provide some semblence of average regional farm earnings before interest and tax (EBIT). For this interest has been added back and the drawings of the owner operator deducted from the MAF model's profit before tax figure.

Farm working expenses should be noted.

As should farm asset values which are after a reduction of $1.22 in the value of Fonterra shares. Despite that, the asset gain over 2008 seems remarkable in relation to the EBIT and gross income figures for some regions.

Ability to Service Debt
Here we reflect on dairy farm ability to service debt at payouts of $4.50, $5.50 and $6.50. The farm is our Waikato average with 2009 costs. Debt has the distribution as in Table 1 and Figures 2 and 3. Levels of production are either typical or optimal for the payout being considered. Interest is calculated at 9.6% - the long term averge interest rate for agriculture according to the National Bank.

It is worth stressing that as payout drops, higher marginal cost production becomes uneconomic and should be dropped. The difference between optimal and typical levels of production is not large at a payout of $6.50, but is still important. As payouts lower the difference becomes critical. The Marginal Cow III provides a full explanation.

It may appear perverse that a lower level of production at lower payouts increases income/reduces losses while increasing the level of debt per Kg of milk solids production. The size of the debt hasn't changed, just the level of production to a more profitable lower level.

The picture painted by Figures 5, 6 and 7 is severe in terms of dairy farm ability to service debt. This situation can only have arisen supported by high and ongoing gains in farm asset values - a classic property bubble or Ponzi scheme. Highly indebted dairy production simply can't service the debt it has, even at high payouts.

Valuation of Assets
The value of any business (including farm) asset is dependent on the stream of income it produces and the rate of tax paid. Before consideraing the value of farm assets it is worth considering the valuation of a pure stream of income. Figures 12 and 13 illustrate the value now of an asset (income stream) - with and without tax - over 20 years at a range of discount rates (effectively the prevailing interest rate) from 4% to 12%. An analogy to the production of one Kg of milk solids production is intended.

The use of income as the basis for the valuation of farm properties is a harsh reality that agriculture has not faced since the reforms of the 1980's. Then though, increasing production was still a viable proposition, credit was available albeit at high cost, and world demand still growing.

Value of Dairy Farm Assets
Figures 10 to 13 illustrate farm asset values based on their income. These are present values for our average Waikato dairy farm over a range of payouts, interest rates and farm costs for 2007 and 2009. Present values are calculated for a 20 year period but take no account of tax on income.

Figure 10 shows that the effect of cost increases between 2007 and 2009 has been to reduce the farm present value per ha by close to 20% ($9,700) at a payout of $9.00 per Kg milk solids, and 40% ($5,800) at $5.00 per Kg milk solids. Figure 11, the comparison on a value per Kg milk solids production differs slightly because at 2009 costs the level of production is reduced from that optimal at 2007 costs.

Both per ha and per kilogram of production values are seriously threatened by increasing costs.

Figures 12 and 13 show the impact of the discount rate (interest rates) on farm asset values. Lower rates significantly increase asset values. Note though that the accepted nominal farm asset values for the Waikato are appropriate for a combination of 4% interest and $9.00 payout. Realistic values for payout and interest rates suggest values closer to 25% of accepted asset values.

Farm Solvency
Figure 14 plots the shortfall in farm asset values against debt for our average Waikato dairy and payouts, but with production scaled to a national level. Debt is distributed as in table 1 and Figures 1 and 2.
Farm assets are calculated as a combination of either the present value of income or nominal asset value per Kg milk solids production plus the value of the farm's Fonterra shares. The discount rate used to calculate present values is 9.6%. The period used is 20 years. Production levels used are optimal and typical. The security value of Fonterra shares (currently $5.57) is considered to be 60% of that price. The nominal asset values (excluding shares) used are $30 and $50 per Kg of milk solids production.

Dairy farm lending makes up 61.5% of bank lending to agriculture while the balance includes horticulture, forestry and sheep-beef operations. Food manufacturing debt is not included in RBNZ figures for agriculture.
This article estimates 91% of dairy farm debt financing is with commercial banks. Some of the remainder is with finance companies.

The distribution of indebtedness between the different financial institutions is pertinent.

Conclusions regards NZ's Agricultural Debt

1. The appreciation of dairy asset values since at least the announcement of what was to become Fonterra has not been based on income and has or is about to stop, and will then go into reverse. Farm values may eventually deflate far enough to reasonably reflect their ability to generate income.
2. Dairy payouts will not improve into or through a global recession, and are more likely to track below long term trends. In all probability the recession will be a minimum of 2, but likely longer than 5, years.
3. The prospects of NZ increasing the intensity of commodity milk production at internationally competitive prices are effectively nil. Retaining international commodity price competitiveness will almost certainly require reducing high marginal cost production.
4. Any significant reduction in milk production will see Fonterra face major capital issues from share redemptions. Fonterra is in a poor position to respond.
5. It is unlikely any dairy farm with debt servicing costs in excess of $3.00 per Kg of milk solids production will survive what is likely to be an extended period of low to moderate payouts. This will not be so much from dairy farms making a loss - which has not been a major issue to date - but more significantly simply from the cessation of asset appreciation.
6. Actual deflation of farm asset values to those reflecting farm income will seriously threaten the solvency of financial institutions heavily exposed to agriculture. Finance companies so exposed may be the first casualties.
7. The survival of dairy farms with more moderate debt ($1.00 - $3.00 debt servicing costs per Kg milk solids production) will depend on:

   a) Ownership structures - owner operator units will have distinct advantages over corporates
   b) Payout and its constancy
   c) Milk production levels nationally responding to international demand, and individually to the expected payout
   d) Costs, particularly non tradable, being contained
   d) A major shift of the dairy industry towards a profit rather than speculative asset appreciation basis. This is a major topic in its own right and will be addressed in another article.

The possible collapse of farm values to those reflecting income will not be the result of the current credit crisis. Rather, the long term inflation of farm asset values has more likely been a contributor to that crisis.

Acknowledgements:
Barrie Ridler of GSL for producing the raw data from which optimal levels of production are derived.

References:
Reserve Bank Information:
Agricultural Sector Indebtedness. Includes debt distribution: Financial Stability Report May 2008

Accurate breakdown of Agricultural Sector Debt as at 30/06/08: Financial Stability Report November 2008

Agricultural Sector Indebtedness. Data files: Data - Financial Stability Report May 2008

Registered Banks Monthly Statistics: Data - Sectoral Analysis of Funding and Claims

MAF Policy Outlook
Check Table 2 - Gross Agricultual Revenue and Expenditure from: Situation and Outlook for NZ Agriculture and Forestry, 2008

Dairy Section: Situation and Outlook for NZ Agriculture and Forestry, 2008

MAF Policy on agricultural debt:
MAF's 1998 Survey of Farm Debt with comparisons to their 1988: Survey National Survey of New Zealand Agricultural Sector Debt 1998

PDF of MAF's 1998 Survey of Farm Debt with comparisons to their 1988 Survey: National Survey of New Zealand Agricultural Sector Debt 1998"

Farm Present Values and long term interest rates: National Bank Rural Report September 2008