The reality of net capital gains and annual profit on NZ primary producing businesses: data from a recent survey of all farm types

ABSTRACT

Annual net cash returns and net capital gains provide farmers with their actual and potential monetary rewards, respectively. This article reports on a national farm survey showing both actual and potential returns, on average, are not as high as might be imagined. This is supported by the findings of earlier studies and has implications for rural societies and their associated economies. Data on income, expenses and profit is presented as is data on farm value changes over the life of each farm. The information is adjusted for inflation to provide ‘real’ data. The profit and net capital gain figures are divided into groupings based on a range of criteria such as total farm capital, and labour compliment as a reflection of farm size. Statistically significant differences occur for some groupings such as farm type and farm size.

KEYWORDS:

• New Zealand
• farm profit
• net capital gains
• survey all farm types
• low returns

Introduction

Everyone involved in primary production and rural society including researchers, management professionals, potential, and existing, investors as well as policy-makers should be keenly interested in the trends in primary production profitability. This article reports on the results of a farm survey which obtained data on contemporary net capital ‘gains’ and returns.

This data must be viewed with the understanding that over time farmers need to become more efficient just to cover the cost price squeeze which is a constant feature of contemporary agriculture and horticulture (Rauniyar and Parker 1998; National Business Review 2017). Furthermore, since 1978, for example, the general inflation rate has meant a 1978 dollar, for example, required $6.20 to purchase the same goods in 2015 (Reserve Bank of NZ 2017). This is a 5.1% annual compound rate over these particular years. This compounding process must be allowed for in any analysis of trends using the rates for the specific years covered. Furthermore, the economic reforms of the late 1980s and early 1990s (Roper 2006) meant that farmers had to rapidly modify their systems to ensure economic survival. This also impacted on land values.

It might be thought land has been a good investment for it is often said farmers’ capital gains offset the lower return on capital experienced relative to returns in other urban industries and investments. Is this true? Data on some land values might suggest this is the case. For example, dairy land in 1978 cost, on average, $13,000 per hectare, and in 2014 the average ‘value’ was $51,000 per hectare (Reserve Bank of NZ 2016), though this has subsequently declined.

The data collected, analysed and reported in this article provides food for thought and debate over the overall profitability of farming, and the consequent future of family farms (Nuthall and Old 2017) which have traditionally been the cornerstone of rural society. The report proceeds by presenting information on farm incomes and costs followed by information on the net capital gains.

Method: information gathering and analysis

In June 2015, a postal survey was sent out to all regions in the country. Approximately 2300 farmers were randomly selected from the national register of farms provided they were full-time operators. The sample was stratified by farm type, farm size (has.) and region. Over 400 replies were received after a reminder letter and a second posting of the questionnaire which was eight pages long covering asset, debt, income, expense and many production and personality issues. After allowing for survey returns from such factors as the death of the farmer the effective response rate was 19%. More responses may have been helpful, but given the detailed and sensitive financial and value information requested the return rate was considered adequate given the levels of variability.

The number of farmers responding, when their farms were classified by size and type, in each group was found to be statistically the same as the national distributions kept by AssureQuality for biosecurity purposes (every farm in NZ is registered with this organisation) from which the sample was selected. (The Wilcoxon Signed Ranks test showed the differences between the population numbers (after adjusting for totals in each group according to the response rate) and sample group numbers were not significant (p = .223), and similarly for the Marginal Homogeneity test (p = .906).) The sample as far as type and size was reasonable. The 2300 farms were 6.4038% of the total population of 35,916 deemed to be full-time operations at the time. The number of registered ‘farms’ is greater than this but many are smaller than an economic unit. The last column of Table 1 gives the number of each type of farm included in the analysis, and the percentage of the total number of each type in the potential sample of 2300.

Table 1. Means for a range of percentage and per labour unit income and expense categories by farm type.

Farm type Income/expense category	Off farm income as % of total net farm income	Farm cash surplus incr. % (ave. last 5 years)	Gross farm income per labour unit $	Farm working expenses per labour unit $	No. of farms and % of sampled popn. farms in brackets*
All farms	24.5	4.9	287,410	158,490	405 (19)
Intensive sheep	16.4	3.8	246,340	123,320	94 (46)
Extensive sheep	17.1	2.8	296,760	171,340	48 (8)
Deer	31.0	5.4	401,430	161,100	10 (19)
Beef	38.8	5.9	207,220	98,210	65 (21)
Dairy	21.2	4.8	335,320	224,890	121 (17)
Other animal	16.2	7.3	98,890	60,330	4 (40)
Fruit	41.5	5.6	370,330	71,620	16 (17)
Arable	26.4	7.5	596,500	310,830	11 (26)
Ornamental/flowers	58.2	7.0	76,670	55,670	5 (11)
Vegetable	27.5	3.7	433,330	393,330	4 (9)
Dairy support	32.9	8.8	186,910	72,710	20 (23)
Other	12.0	7.4	299,370	64,200	7 (29)
F test sign. probability	.001	.658	.000	.035

*Further details of the sampling and sample are given in Greig et al. (2017).

The response questionnaire information was entered into a database (SPSS, Privitera 2014) to allow analyses covering a wide range of issues one of which was the profitability of farming. The questions were all answered by the farm managers/owners so the accuracy will be reasonable. Furthermore, the responses were all voluntary so there is no reason to believe there was any bias particularly as the objectives were pure research rather than attached to any vested interest situations. Various studies, for example, Rangul et al. (2008), have shown self-reporting is relatively accurate. Anonymity was guaranteed and the questionnaire vetted by the Lincoln University social science human ethics committee.

Income situation

Tables 1 and 2 provide data on farm income, and some expense items. Some of this information is presented on a per labour unit basis to compensate for the different farm types which on a per hectare basis are not comparable due to soil quality and climate variations. On the other hand, percentage figures are directly comparable. Labour unit numbers include the manager.

Table 2. Per farm percentage return on capital for a series of categories.

Farm type	Mean % return on capital	Total farm capital (million $s)	Mean % return on capital	Age of farmer (years)	Mean % return on capital	Highest education level (and years of education)	Mean % return on capital
All farms	2.5		2.5		2.5		2.5
Int. sheep	2.8	<2	1.4	<45 years	2.7	1°	–*
Ext. sheep	1.9	2–4	2.0	46–55	2.1	2°< 4	2.7
Deer	4.5	4–6	3.4	56–65	3.3	2°> 3	3.4
Beef	4.2	6–8	4.8	>65 years	1.8	3°< 3	1.5
Dairy	2.1	8–10	1.9			3°> 2	2.2
Other animal	–^a	10–12	1.6
Fruit	0.2	12–20	1.7
Arable	0.7	>20	2.1
Vegetable	11.7
Grazing	2.6
Other	−1.4
F test probability	.027		.165		.266		.178

^aAn insufficient number of farmers provided data in these categories to warrant calculation.

The differences in ‘off farm’ income, gross income per labour unit and working expenses are all statistically different across farm types. On the other hand, the farm cash surplus change per year is not statistically different. In this latter case, it will be noted the average for each farm type hovers around 5% and averages 4.95% across all farms. What is interesting, as noted in the introduction, is the CPI indicates an annual compound change of 5.1%. Up to 2015, the cash surplus is barely holding its own. The best performers in this survey were arable, dairy support and some horticultural operations. ‘Other animal’ is also high, but the numbers of this class in the survey were low.

It is also interesting to see ‘off farm’ income was surprisingly high with an average across all farms of nearly 25% of net income. In an increasingly uncertain world, this diversification is sound. Of broad acre farming, beef farmers, and deer like to diversify ‘off farm’ more than the others.

The high gross income per labour unit for deer farming reflects in part the low labour requirements. Some of the horticultural occupations similarly provide high figures, as does dairy. Expense wise, arable is high as is vegetable production.

Across all farms there was a $128,900 surplus per labour unit (including the manager) to cover other expenses such as interest, depreciation and lease/rent items.

It is interesting to note that Eves and Painter (2008) found over 1990–2005 net income per farm in NZ increased 5.4% per year compared with the 4.9% found in this study over a shorter, and more recent, period.

Turning to Table 2, the ‘all farms’ return on capital is 2.5%. This is certainly not high and reflects the overall situation for 2014/2015 given the data was collected in late 2015. The highest return is from beef followed some way down by extensive sheep. The high dairy land prices are reflected in the 2.1% return. It will be noted the farm type differences are significantly different. But this is not the case for the differences across farm size re the investment, nor across farmer age or her/his education level. That is, TFC (total farm capital), farmer age and education levels do not seem to provide a large impact on the return on capital. For earlier periods the Dairy NZ Economic survey (range of years), which comprises volunteering farmers, estimated an average return on assets over 1998 to 2016 as 7.7%, and the Beef and Lamb NZ Survey (range of years) data from randomly selected beef and sheep stratified farms over 2008–2017 provides an equivalent figure of 1.0%.

Linking the survey ‘all farms’ average of 2.5% to the range for this statistic gives a standard deviation of 3.16% and data showing 14.9% if farms achieved less than 0%, 14.4% less than a positive 1% return, between 1% and 2% were 2.2% of the farms, 13.9% between 2% and 3%, 11.7% between 3% and 4%, and 18.1% had a return on capital above 4%. As would be expected, some farmers are producing an acceptable return, but the majority are well below urban expectations.

Net capital gains across the years

The farmers were asked to provide the year of purchase of their farm, the purchase price, current valuations for 2010 and 2015 as well as the level of debt at each date. This information enables calculating their net capital gains through assessing their asset, debt and equity changes over the life of the farm.

This base data is given in Table 3. Also provided is the data relative to TFC levels, debt levels and labour units.

Table 3. Per year percentage change in assets, debt and equity from purchase of farm to 2015 by farm type, Total Farm Capital, labour unit groups and debt groups. Nominal figures.

Farm type	Asset change (%)	Debt change (%)^a	Equity change (%)	Debt group ($0000s)	Asset change (%)	Debt change (%)	Equity change (%)
Int. sheep	3.52	2.33	3.67	<$25	2.77	−25.66	3.09
Ext. sheep	3.08	−2.08	3.48	25–50	3.41	−2.20	3.73
Deer	2.67	−5.46	2.82	50–75	3.00	1.50	3.25
Beef	2.71	−18.89	3.03	75–100	3.37	2.71	3.57
Dairy	3.62	3.01	3.37	100–200	3.81	2.88	4.20
Other animal	3.52	1.98	4.27	200–300	4.33	4.18	4.37
Fruit/viticult.	3.34	4.25	3.50	300–500	4.09	3.69	−0.92
Arable	2.70	−12.59	2.79	500–800	3.01	4.59	4.46
Nursery/flrs.	3.03	NA^b	3.23	>$800	3.50	3.38	3.82
Vegetables	3.82	NA^b	3.94
Grazing	3.05	−6.80	3.16
Other	3.06	−15.00	3.89
All farms ave	3.30	−1.71	3.38
F Sign. prob.	.916	.110	1.000	F Sign. prob.	.348	.002	.084
Total farm capital				Labour units
<2 million	3.19	−37.69	3.51	≤1	3.20	−9.29	2.37
2–4 million	3.11	1.38	3.40	1–2	3.25	−3.52	3.56
4–6 million	3.15	−0.83	1.86	2–3	3.90	3.05	4.18
6–8 million	3.46	2.68	3.70	3–4	2.78	−0.33	4.28
8–10 million	3.40	0.78	4.22	4–5	3.15	3.07	3.28
10–12 million	1.72	3.24	4.07	5–6	4.44	4.45	4.46
12–20 million	3.76	1.39	3.81	7–9	3.44	3.10	3.94
>20 million	3.47	3.22	3.78	>9	3.28	3.12	3.39
Sign prob. (F)	.286	.000	.437	Sign. prob. (F)	.794	.643	.590

^aNote that a negative means debt has INCREASED.

^bNA = Not available.

It is only the debt change levels that approach statistical significance across different farm types, different debt level groups and TFC groups (as you would expect). It will be noted the asset value change across all farm types is just over 3% with the debt change increasing slightly leaving just over 3% change to the all important equity values. Within these values, there are some interesting figures such as the debt changes for beef, arable and ‘other’ farm types. Some of these figures will likely to be due to chance both over sampling and which farmers actually chose to answer the survey.

It is these nominal figures (Table 3) which farmers, and many others in non-rural societies, probably take note of without allowing for the inflation rates. By introducing inflation through compounding the data to a base of 2016, a different story emerges (Table 4).

Table 4. Per year percentage change in assets, debt and equity from purchase of farm to 2015 by farm type, Total Farm Capital, labour unit groups and debt groups. Real figures with 2016 base.

Farm type	Asset change (%)	Debt change (%)^a	Equity change (%)	Debt group ($0000s)	Asset change (%)	Debt change (%)	Equity change (%)
Int. sheep	−0.48	−0.53	−0.06	<$25	10.41	−0.29	−11.05
Ext. sheep	10.81	−0.67	−11.39	25–50	−0.08	−1.45	−0.44
Deer	0.03	−0.64	−0.68	50–75	0.58	−0.37	−0.95
Beef	0.11	−0.48	−0.40	75–100	−0.35	−0.49	−0.18
Dairy	0.79	−0.10	−0.81	100–200	−0.16	−0.55	−0.13
Other animal	−0.75	−0.68	0.07	200–300	−0.38	−0.45	−0.07
Fruit/viticult.	−0.39	−0.50	−0.11	300–500	−0.43	−0.35	0.10
Arable	0.34	−0.30	−0.64	500–800	1.40	0.14	−1.32
Nursery/flrs.	−0.76	−0.30	0.49	>$800	1.99	0.89	−1.14
Vegetables	−0.67	−0.57	0.10
Grazing	−0.67	−0.55	0.07
Other	−0.63	−0.61	0.02
All farms ave.	1.37	−0.38	−1.68
F Sign prob.	.795	.947	.768	F Sign. prob.	.647	.000	.617
Total farm capital				Labour units
<2 million	−0.50	−0.46	0.03	≤1	−0.51	−0.51	0.02
2–4 million	5.04	−0.41	−5.50	1–2	3.47	−0.69	−3.99
4–6 million	−0.14	−1.09	−0.42	2–3	−0.07	−0.38	−0.31
6–8 million	−0.23	−0.53	−0.29	3–4	−0.45	−0.43	−0.03
8–10 million	−0.46	−0.48	−0.02	4–5	0.04	−0.23	−0.29
10–12 million	0.06	−0.36	−0.41	5–6	−0.56	−0.29	0.27
12–20 million	1.17	−0.02	−1.28	7–9	4.50	1.52	−2.98
>20 million	2.09	0.87	−1.26	>9	1.84	0.60	−1.31
F Sign prob.	.892	.001	.893	F Sign prob.	.951	.001	.960

^aNote that a negative means debt has increased.

As the farms were purchased over a wide range of years it should be noted the data in Tables 3 and 4 have been converted to a ‘per year’ basis in contrast to a ‘life of the farm’ basis. This provides comparability.

Again, across farm types, TFC, debt groups and labour units there are significant differences for the annual debt change figures. But this is not the case for asset and equity figures. With this in mind, the ‘all farms’ averages show real net capital gains have been close to zero with the net asset change slightly positive. Most farmers would have liked this latter figure to be much higher. It might have been thought farm size would influence these outcomes (TFC & labour units), but this is not the case. On the positive side, it can be concluded farmers, in recent times, have not bid too high for land in that the real values are virtually static.

While, on average, net capital gains have not been positive for farmers, this is not the case for some individuals. The standard deviation on the equity change percentage per year was 20.8 showing the considerable variability. On the positive gains side of the distribution, 58.6% were over 0%. This is balanced by 14.7% lower than −1.0%.

Some farmers certainly maintain they have made good real capital gains (NZIER 2014). This will undoubtedly be the case for some depending on purchase dates, farm type and region. To explore one possibility, only farms purchased after 1980 were included in the analysis. However, this showed the ‘all farms’ real asset change was −0.61% per year, real debt change −0.47% per year and the equity change 0.16% per year. When comparing farm types the changes were not significantly different, but when the farms were grouped according to debt levels the asset and debt changes were significantly different across increasing debt categories. The same occurred for groupings according to Total Farm Capital, but for groupings across labour input levels it was only the percentage change in real assets that were significantly different.

The overall story has not changed much for this particular set of years (1980 or later purchase) relative to the base ‘all years’ data. The ‘all farms’ averages, while slightly different, provide much the same picture. The year chosen (1980) was arbitrary. No doubt selecting all farms purchased after, say, 2000, would have provided different figures. When 1990 was used the ‘all farms’ average per year net equity change was 0.23%. This is slightly higher than the increase in real values than the 1980 cutoff year. This is relative to the approximately 9% per year land price increase found by Harrison and Hoskin (2011) for the years 1990–2010 using Quotable Value (QV) data. And Dunston and McDonald (2012) calculated from QV rural indices that the index increased from 1 to 2.5 over 1997–2011 (10% per year). These figures are reinforced by Eves and Painter (2008) who give 10.7% average real growth of land value over 1990–2005.

As noted, individual farms would have experienced different net capital gains, some very favourable, others not so. A case in point is given by Pangborn (2012) who reports on a case study of a dairy farming enterprise in Canterbury started in 1988 and analysed through to 2010. The compound real gain was reported to be 18.9%. If this analysis was repeated through to 2018 the return would have been different given dairy land values have now declined.

In considering the equity figure changes it is interesting to note in 1990 the Reserve Bank (2018) reports a total farm national debt of $5,014,000,000 which increased to $60,529,000,000 by 2017. The rate of increase has varied from −1.1% in October and November 2011 through to a high of 25% in May 2002. While real land values might have increased markedly in some cases, the increasing debt levels mean the net equity gains have been minimal as the survey data shows. For example, using QV NZ data, Stillman (2005) found the real value of all land, including both rural and urban, averaged an increase of 16.6% per year over 1989–2003. In contrast, the Reserve Bank (2016) found average dairy land increased 4.5% per year over 1978–2014. As the years of ownership of the farms included in the survey vary markedly, it is not possible to make perfect value and equity change comparisons from the literature as most reports cover quite different eras. The outcomes vary with each particular set of years. The survey farmers’ ‘years of ownership’ covers many decades through to a few years better reflecting the longer term situation.

Discussion

While the figures presented are from a sample in contrast to all farms, the representative nature of the sample suggests the data provides a reasonable indication of the current annual profit and net real capital gains situation for a significant section of the rural community.

Given the low level of annual return (2.5% on capital), and virtually zero net real capital gains, it is clear farmers, and their families, must obtain many side benefits from farm life compensating for the low returns. Rural living seemingly does suit many families. The real issue is whether they cover their cash costs and have sufficient income left for a reasonable way of life (Gasson 1974). Considering only cash income and outgoings, the net cash return per labour unit of the farmers providing this detailed information was $73,526 for 2014. There was considerable variability in this figure particularly depending on debt levels (29.4% of the properties were debt free by 2015) with 14% showing a cash surplus of $200,000 or greater, but on the other hand 8.1% returned a cash loss.

Despite this, some individual farmers and investors will have made real net capital gains, and possibly also high annual returns, if, more than likely by chance, they purchased at a low point in land values, and then sold at a high point, and similarly for product prices and annual returns. Very few would be able to successfully predict the low and high points.

In assessing the data presented it must be remembered while land values have shown real increases, the concentration in this work has been on the real net equity gains. It has been shown overall the national farm debt has increased markedly impacting on real net gains. It is also relevant to consider the factors which drive land price changes. Allan and Kerr (2014) found profit, bank credit policies and amenity values were significant drivers, but also commented they thought there were irrationalities in the process inherently exhibited by farmers. In addition, Just and Miranski (1993) studied the US farm land market and concluded the drivers, as might be expected, were real returns, inflation, current land price swings and farmer expectations.

The farmers in this survey were all practicing farmers. Clearly any who had left the industry and consequently had actual sale figures do not appear. The reasonably high level of ‘off farm’ income must also be born in mind when assessing the significance of the farm incomes. And of course, most farmers and their families do not actually realise any capital gains as the majority of farms are passed onto children in contrast to being sold to others. However, rising net worth does allow additional borrowing for development and other non-farm ventures even if any gains are not directly cashed up.

It is also relevant to consider previous surveys. A 2013 survey showed farmers reported data giving a 3% return on capital (Nuthall and Old 2014). This also complies with Nartea and Webster (2008) who, for an earlier period (1988–2003), found farms on average returned a profit of 2.76%. However, they also found nominal asset values increased 9.83% for this period. Furthermore, Johnston and Frengley (1994) report, when reviewing the viability of farming, a real net worth index for all farms of 1590 in 1984, 740 in 1987, 910 in 1990 and 1000 in 1993. Over this decade net worth had declined by a third. For the same years, they also report a return on capital of 2.7%, 5.4%, 5.4% and 4.0%, respectively, with cash savings per farm of $8575, −$715, $969 and −$4139 which highlights the nature of primary production. Also, Allan and Kerr (2014), in assessing the drivers of rural land value, provide a graph of rural sale price from 1980 to 2012 and found the increase was approximately an average of 1.8% per year. Despite the reports all taking different time spans, the current situation seems, on average, to be a replica of the recent, and not so recent, past. And it is likely this situation has existed over many eras. Greasley and Oxley (2009) present a graph showing a land price index set at 100 in 1913 dropping to 76 by 1939.

If farmers wish to make real capital gains they will need to initially purchase at lower prices in keeping with conservative estimates of future product prices and increasing costs. It is possible, for example, the availability of the farm ‘next door’ tempts farmers into paying too high a price to allow making an annual return on capital similar to urban-based fixed interest investments (currently around 4.3% for five years). However, given the non-monetary benefits from primary production (Gasson 1974; Nuthall 2009) farmers are most unlikely to become purely economically rational. There is no reason to believe this situation is likely to change in the foreseeable future leaving rural societies to carry on much as they are today except perhaps with smaller populations as farm’s amalgamate (Hooper et al. 2002).

Acknowledgements

We gratefully acknowledge Lincoln University and all the farmers who took the time to complete the questionnaire.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

Lincoln University provided the funds for this research.