His theory may now be able to be quantified as Dairy New Zealand, (to which all NZ dairy farmers including Craigmore contribute) have recently estimated returns on owner-operator skill in NZ dairy farming. It turns out that, across the nearly 12,000 dairy farmers in New Zealand, top quartile farmers produce profits of [NZ$3,000] per hectare per year.  This is 50% more than the average dairy farmer who generates NZ$ 2,000.[ii]

Distribution of owners-operators’ operating profit NZ$/ha

Source: DairyNZ Economic Survey 2009-10

This study was carried out for the farming financial year ending June 2010, when NZ dairy farmers faced broadly the same climatic conditions and commodity markets across the country. The authors corrected regional differences of soil type, altitude and moisture. Based on this study, and assuming these differences persist from year to year under management of a particular farmer, it seems correct that, in farming, there is indeed a wide dispersion of returns depending on farmer skill.[iii]

What are the implications for investors in farmland? We have argued elsewhere that farmland is a cash-positive, inflation-proof store of value with bright prospects given changing tastes in Asia. But is this “macro-story” enough? If one wants to own farmland how does one execute on this view? How does one, or one’s agent or fund-manager, actually own and oversee farmland investments?

One logical way to invest in farmland: become a “farm landlord”

One logical solution to the “operational risk” of farmland is to invest as a landlord and lease out the farm. To become, like the aristocrats of old, a landlord to a series of tenant farmers. These tenants will compete to pay the highest rental payments i.e. they will “self select” to bring (at least some of) the benefits of “top quartile” productivity to the investor.

In certain sectors (e.g. straightforward farming operations where the contracting costs of protecting asset quality are not too high) and at certain times (such as now when rental yields on land in New Zealand are at the high end of their historic 2% to 4% range) this is an attractive strategy.

Farm-landlord vs. commercial property comparison

Why is this attractive when commercial property in NZ is currently yielding 7%? The reason is that farmland has very low depreciation, relative to commercial property where the building must be replaced every 25 years (depreciation of around 2.5% p.a.). Further farmland is estimated by the FAO to face 2% to 2.5% annual terms of trade improvements during this decade. We calculate that, therefore, despite a lower headline yield, a landlord strategy should out-perform commercial property by around 2% p.a.

Farm-landlord vs. share-farming

Landlord systems seem very logical. One attraction of the “farm-landlord” approach is that returns should have low volatility. The tenant should absorb annual returns volatility from weather-related yield and commodity price fluctuations while the landlord (provided the tenant remains solvent and leases are rolled over at reasonable levels at the end of rent periods) enjoys a constant monthly rental.

However farm-landlord ownership of farmland is not very frequent. 90% of farms in the USA, for example, are managed by farmers with partial or full ownership of farm equity[iv]. Why is this? And why are there not more “corporate farmers” (e.g. managers of multiple farms listed on stock markets)?

There are two main reasons why the owner-operator approach predominates in farmland. Firstly, owner-operators accrue productivity and development gains (enhancements to the value of their farms), whereas tenants (and corporate managers) seldom have incentives to add such value to their landlords (and shareholders). Whilst landlords and managers can contract and attempt to enforce negative requirements (e.g. tenants maintain fertiliser application and therefore soil fertility), it is difficult to contract/enforce positive gains. A tenant is less likely than an owner-operator to invest appropriately in the soil, pasture, irrigation, stock-water, drainage, facilities, lanes and other infrastructure, technology, plant-genetic and other improvements, all of which lift the productivity of farms.

This is a problem for farmland landlords as the benefits of productivity growth and development gains are significant. A system of ownership which provides no incentives to productivity improvement is “leaving a lot of money on the table”. To illustrate: average NZ dairy farm per hectare productivity has historically grown by 2% per annum (from production of 650 kg of milk solids per ha in 1992 to over 900 kg of solids by 2010).

These gains have been achieved in a sector that is either entirely owner-operated, or share-milked, but where in either case there is a high level of owner-input into productivity improvement decisions.  Further, we estimate development gains (change of crop-type) have added a further 1.5% or 2% per annum. These productivity and development gains will be, naturally, capitalised into land valuations since land is valued as a multiple of its expected cash flows.

Land-use changes (i.e. “development” returns) are, again, more available to the owner-operator than to the passive owner. It is possible in New Zealand, for example, to take irrigated mixed sheep and arable operations yielding revenues of NZ$ 3,000 per hectare and to convert this to dairy farming yielding revenues of NZ$ 8,000 per hectare. A landlord and tenant team might contract to capture these gains and manage the risks, but it is rarely seen given the risk and long time-horizons of such developments.

In the Craigmore Farming Partnership all farm managers are equity partners in their farming businesses (subject to satisfactory trial periods). Going forward we estimate productivity growth of 1.5% p.a. and averaged development gains of 1.5% p.a. will drive at least 3% per annum of annual returns at the asset level. In fact on many of our farms we estimate 5% p.a. in the early years after acquisition.

The second reason that owner-operators out-perform landlords is that operational farming is more profitable than “land-lording” in cash-flow terms. As discussed above, the operating farmer is exposed to the annual volatility of commodity prices and of climatic fluctuations. The superior returns of operating farming are a reward for absorbing this volatility, as well as a return on skill. Working farms in New Zealand can presently be purchased at cash-flow yields of over 6% (using 5 year normalised commodity prices). This in contrast to farmland rental yields of 4% in NZ at present.

In the table below we contrast “owner-operator returns with our estimates of “fair” commercial property and farmland-landlord investment returns.

We subtract management and governance overheads. These are assumed, before leverage, at 0.75% for commercial property, 1% p.a. for farm-landlords and 1.25% where a fund partners with share-farmers on each farm.

We estimate that the share-farming approach should generate 3.5% p.a. more than the farmland-landlord approach, on average. Note that this analysis is asset-level only and does not review the impacts of leverage, taxes or changes in the valuation multiples of land.

We believe these additional returns are sufficient to compensate for greater short-term returns volatility and operational risk of an active farming portfolio, i.e. larger requirement for oversight and expertise.

Furthermore, many of the key production risks can be effectively managed through sound investment and management decisions. For example, variable rainfall often poses the greatest threat to farmland production. Therefore selection of irrigated or ‘reliable rainfall’ properties significantly reduces the operational risk taken on by investors.

“Transport costs matter”

Distances from the consumer are definitive drivers of value (and choice of crop type) for farmland investments. It is not economic to grow bulk grains in the middle of a continent like Brazil, for example, and then try to transport them to international markets. The cost of transporting them out would be greater than their value (for this reason Brazil is shifting populations into land-locked areas 2,000 km from the sea-lanes like Mato Grosso, to bring industry and consumers closer to the sources of raw materials). These middle areas of Brazil, on the other hand, are terrific places for growing beef which is a much higher value product per kg (by a factor of over 10), so transport costs are a much smaller fraction of beef commodity prices.

Craigmore Research invites you to join us in a discussion about quantifying and exploring these cost drivers of farmland profitability, crop choices and land values at two seminars in the UK in late June.

An invitation is attached below for a lunch seminar on this topic on Wednesday 27th June. Please drop me an email if you would like to attend.

Best regards,


[i] He also used to say that the best practitioners were normally 5 to 10 years ahead of the universities in their techniques.  That is a story for another Commentary.

[ii] Converting this average dairy farmer’s profits into US$, and converting the metric from hectares to acres.  This represents profits of [US$ 700] per acre per year so still not bad!

[iii] The data needs to be treated with some care. Returns per hectare are not the same as returns per $ of capital employed since some farms are more expensive, on a per hectare basis, than others. Also some of the farms in the population are likely to have been “under development” i.e. recent conversions from sheep to dairy farming where the pastures will not reach full productivity for 2 to 4 years. These factors mean there could be some “noise” in the data. Craigmore Research is planning to re-visit the data to attempt to correct for these biases.
[iv] 4USDA 2007 Census of Agriculture.


Published: 1 June 2012