Information only disclaimer. The information and commentary in this email are provided for general information purposes only. We recommend the recipients seek financial advice about their circumstances from their adviser before making any financial or investment decision or taking any action.

For a long time, one of the most successful ways to build wealth in New Zealand farming has been relatively straightforward: run the farm well, build equity, borrow some more and buy more land.

There is nothing to suggest that model is finished. A well-bought dairy farm that generates enough cash to service its purchase price can still be a very good investment, as can a horticulture conversion that materially lifts the productive capacity of land. Scale creates efficiencies, and good land remains scarce.

But there is another question worth asking before buying the neighbour: What return could I generate by putting that same capital into the farm I already own?

I think that question will become increasingly important over the next decade.

We’ve already been doing it

We don’t have to imagine what productivity-led growth might look like because New Zealand farming has been doing it for decades.

In 1992/93, the average New Zealand dairy farm produced 653kg of milksolids per effective hectare and 259kg per cow. By 2024/25 those numbers had reached 1,137kg per hectare and 414kg per cow – increases of around 74% and 60% respectively.

Source: DairyNZ, New Zealand Dairy Statistics 2024/25.

The sheep story is just as impressive. Since 1995, lambs produced per adult ewe mated have increased from 100% to 137%, while the dry matter required to produce a kilogram of carcase weight has fallen from 36kg to 24kg. Put those gains together with heavier carcases and the result is striking: carcase weight sold per ewe has doubled in 30 years.

Beef + Lamb New Zealand attributes around half of the efficiency improvement to genetic progress, with better management helping increasingly capable genetics to perform.

Source: Beef + Lamb New Zealand, 30-year sheep industry report card.

These are extraordinary productivity gains, but they also illustrate something important for the next phase of farming. Wealth creation hasn’t come only from owning more hectares; a significant amount has come from making the hectares, animals and capital we already own work harder.

Finding the hectares we already own

Virtual fencing may provide one of the clearest examples of what productivity investment could look like over the next decade, particularly on hill country.

Pamu’s 2,879ha Eweburn Station near Te Anau is already using Halter virtual fencing to control and shift cattle without conventional subdivision. More than 80% of the pasture is now being accessed, including older pasture that cattle can be directed onto through virtual breaks.

The bigger opportunity, though, may be what happens when virtual fencing changes the grazing system itself.

Former Beef + Lamb New Zealand chair and farm consultant James Parsons has been using the technology on his 600ha Ashgrove Farm in Northland, around 70% of which is hard hill country. Instead of leaving cattle in large paddocks for a week or more, virtual boundaries allow them to be shifted much more frequently and grazing pressure directed onto areas that previously tended to be lax grazed.

That starts to bring some of the disciplines of rotational grazing into country where conventional subdivision has historically been too expensive or impractical.

This distinction is important. The value of virtual fencing isn’t simply the saving from not building another physical fence. If it allows a farmer to harvest more of the grass already being grown, improve residuals and regrowth, lift animal performance and make previously underutilised country genuinely productive, the potential value is considerably larger.

AgFirst modelling associated with Parsons’ work illustrates the opportunity. On an extensive hill-country system, it modelled a 21% improvement in pasture utilisation together with an additional tonne of dry matter grown per hectare through improved grazing management by year two. The resulting improvement in gross margin was modelled at $348/ha.

Those are modelled rather than realised results, and the outcome will obviously vary between farms. But the underlying idea is powerful. For generations, one of the constraints on intensifying hill country has been the economics and practicality of subdivision; virtual fencing potentially changes that equation.

In some cases, the cheapest additional hectare may be one you already own but haven’t been able to fully harvest.

Genetics may still have a long way to run

The historical sheep numbers demonstrate how powerful genetics have already been, but selection is increasingly moving beyond simply producing more lamb or faster growth. Increasingly, the question is whether we can breed an animal that produces as much or more while requiring fewer inputs.

Nextera Genetics is an interesting example. Its Commando shedding genetics are being developed around a lower-input sheep system that removes wool harvesting while retaining meat production.

Recent modelling undertaken for Willesden Farms compared its traditional Romney system with a steady-state Commando system. Despite losing the wool income, the modelling showed operating costs falling by around $206,000 a year, with sheep-enterprise profit increasing by approximately $146,000, or 14%. The major gains came from removing shearing and wool-related costs and reducing labour and animal-health requirements.

The wool-price comparison is particularly interesting. The Willesden analysis calculated that wool would need to average $7.66/kg greasy across the whole clip for the Romney system to produce the same operating profit as the Commando system.

An entirely separate AgFirst analysis of a 770ha North Island sheep and beef farm arrived at an almost identical figure of $7.77/kg greasy. That analysis modelled a 19% lift in EBIT from moving to a fully shedding flock, including lower wool-related costs and a 7% increase in sheep-meat production per hectare.

Source: Nextera Genetics

The opportunity isn’t simply a sheep that doesn’t need shearing. It is using genetics to redesign the economics of the entire production system.

The same principle applies in dairy, where genomics, sexed semen and increasingly targeted breeding are accelerating genetic gain. Unlike buying another farm, genetic improvement isn’t a one-off increase in scale; its benefits can continue to compound through the herd or flock.

Sometimes productivity means growing something different

There is another way to increase the earning capacity of a hectare: change what we grow on it.

Kiwi Quinoa is a good New Zealand example. Dan and Jacqui Cottrell began experimenting with quinoa in 2014 after identifying an opportunity for a niche, higher-value crop. Their selected variety suited their growing conditions, was naturally free of the bitter saponins that normally require removal, and had a relatively short growing season. They produced their first commercial crop in 2016.

The business has since moved well beyond the original trial, working with growers across New Zealand and supplying wholegrain quinoa as well as processed products.

They weren’t simply asking how to produce another tonne of a conventional crop. They were asking whether the hectare could produce something worth more.

That distinction matters. There will be other opportunities through new crops, varieties and uses for existing crops, although they come with a different set of risks. Agronomic success doesn’t automatically create commercial success; market depth, contracts, processing and ultimately the customer become just as important as what happens in the paddock.

Sometimes productivity means taking cost out

Kaiwaiwai Dairies in South Wairarapa gives us a very different example.

The 900-cow winter-milking operation installed a 54kW ground-mounted solar system in 2021 at a cost of around $110,000. More than 90% of the electricity it generates is consumed on farm through activities such as milking, milk chilling, irrigation and effluent management.

The economics are what make it interesting. EECA reports annual electricity savings of $17,000-$20,000, with the cost of solar-generated electricity estimated at around 15 cents/kWh compared with 28 cents/kWh for grid power. Its current return on investment is around 17% and approaching 19% at full performance.

Interestingly, selling electricity isn’t the attraction. Despite being grid connected, annual export income is only around $267; the value comes overwhelmingly from replacing electricity the farm would otherwise have bought.

Kaiwaiwai isn’t necessarily producing another litre of milk because it owns solar panels. It has simply deployed $110,000 of capital into an asset that materially lowers the cost of producing the milk it was already going to produce.

Three sources of productivity

It is useful to think about agricultural productivity as coming from three broad sources. Biological productivity – genetics, fertility, pasture management and nutrition. Physical productivity – irrigation, drainage, subdivision, machinery, housing and animal infrastructure. And increasingly, precision productivity – using data, sensors, automation and AI to manage inputs and production at a much finer level.

These aren’t stages where one replaces the other. All three continue to develop, and some of the biggest opportunities will come from combining them. Better genetics still matter enormously, as does getting water, feed, infrastructure and the physical farming system right. Precision technology adds another layer by allowing those biological and physical assets to be managed more accurately and efficiently.

Historically, much of farming has been managed using averages. A paddock receives an application, a mob receives a treatment or a crop is sprayed. Technology increasingly allows us to identify variation within those groups and act on it.

We can already see this in New Zealand. Vantage NZ’s WeedSeeker technology, for example, identifies weeds and activates the spray nozzle where required rather than blanket-spraying the whole area. Vantage says chemical use can be reduced by up to 90% in appropriate applications, while weed mapping records where problems are occurring.

LeaderBrand has taken the concept another step with the introduction of a Carbon Robotics G2 LaserWeeder into its Gisborne vegetable operation. High-resolution cameras scan the crop, artificial intelligence distinguishes weeds from the plants being grown, and lasers kill the weeds without spraying or mechanically disturbing the crop.

LeaderBrand is currently trialling the technology, with performance, cost and return-on-investment data intended to be collected and shared with the wider industry.

John Deere’s See & Spray technology is simalr and demonstrates where this can go at enormous scale. Cameras mounted on the spray boom identify individual weeds and machine-learning systems activate individual nozzles only where required. Deere says that during the 2025 North American season the technology was used across more than five million acres, reducing non-residual herbicide use by nearly 50% on average.

The bigger point isn’t really about weeds. Per paddock is moving towards per plant, per mob towards per animal, and increasingly crop and pasture management towards the square metre.

Over time, the same principle can extend across feed allocation, animal health, fertility, genetics, fertiliser, irrigation and pasture management. Humans could never economically make millions of individual decisions across a farming business, whereas AI and automation increasingly can.

That ability to move farm management from averages towards individual decisions – at the level of the animal, plant or square metre – could become one of the next significant productivity opportunities in agriculture.

Connecting the data

There is another part of this that may ultimately be just as important as the machinery itself.

Farms are already generating enormous amounts of information through soil and yield data, fertiliser and spray records, irrigation, effluent, weather, machinery and animal performance. The problem is that much of it has historically sat in different places.

New Zealand company Precision Farming provides an interesting glimpse of where this could head. Its software connects fertiliser and chemical ordering systems such as MyBallance and HawkEye with machinery platforms including John Deere, Trimble, Topcon and Tirotech. Jobs can move digitally to machinery and proof-of-application data can flow back to the originating farm system for compliance, analysis and future decision-making.

That sounds less spectacular than firing lasers at weeds, but connecting the information may ultimately be just as important. Once these datasets begin talking to one another, the opportunity moves from precision application towards precision decision-making: what happened here, what did we apply, what did it cost, what did it produce, and what should we do differently next time?

Think about the next $1m

Imagine a farming business has another $1m of equity or borrowing capacity available to invest. It could form part of the equity required to buy another property, and that may well be the best investment.

But the same capital could potentially go into genetics, virtual fencing, intensive finishing systems, automation, robotics, precision application, on-farm energy, data and decision-support technology, or fundamentally changing the production system itself.

Some of these investments may look quite different from the assets farmers have traditionally borrowed to buy, but the investment question should be exactly the same: Where does the next dollar of capital produce the greatest sustainable increase in earnings?

That means measuring returns properly – not simply whether an investment increases production, but what happens to sustainable earnings after the additional operating costs and capital required to generate them.

It gives us a common basis for comparing investments that can look completely different, including buying the farm next door.

But can we finance it?

There is a challenge here for those of us involved in providing capital to agriculture.

Our traditional agricultural lending model has generally been very good at financing assets that can be valued and secured. Borrow $5m to buy another farm worth $7m and there is another $7m of land sitting on the balance sheet; the lender can value it, take security over it and readily understand where its money has gone. Invest $1m improving the productivity of the farm you already own and it can be more complicated.

Some improvements – irrigation, drainage and other permanent physical improvements – may be relatively readily reflected in the value of the property. Others are less straightforward. Housing and animal infrastructure can require substantial capital but may only be partly reflected in a subsequent land valuation. Better genetics, technology, automation, software or investments that materially lower the cost of production may create significant additional earnings while adding little traditional security value at all.

A very good investment can still be difficult to finance if the lending decision is driven mainly by what happens to the traditional security value.

For productivity investment to attract capital, farmers will need to present it differently. The investment needs to be fully costed and longer-term budgets need to clearly isolate the marginal earnings it is expected to generate. Importantly, that calculation needs to include the full economic cost – including depreciation and, particularly with technology, the risk of obsolescence or replacement as better technology emerges.

The assumptions behind those returns need to be well supported, with clear milestones and controls so both farmer and capital provider can see whether the investment is delivering what was expected. Where utilisation is critical to the economics, longer-term supply, processing or offtake contracts may also become important. There is little point investing heavily in productive capacity without sufficient confidence that it will be used.

The financing itself will evolve as well. I think it will become increasingly commonplace for finance to arrive with the technology rather than through the farm’s traditional bank. Equipment and technology providers will increasingly package finance with their products, while subscription, lease, usage-based and other service models may allow farmers to access technology without owning the underlying asset at all.

That makes sense. A technology provider may understand the useful life, residual value, upgrade cycle and economics of its product considerably better than a traditional property lender. Financing can then be matched to the economic life and cash returns of the investment rather than simply being added to long-term farm debt.

Traditional lenders will still have an important role, but the structure should reflect the economics. Where an investment generates strong cash returns, a shorter amortisation period may make more sense than treating it like long-term land debt. Capital providers will also need to become better at recognising the enterprise value created by productive investment, even where a traditional property valuation captures only part – or none – of that value.

None of this diminishes the importance of land security or conventional bank lending. It simply means becoming better at funding the earning capacity of the farming enterprise as well as the value of the land underneath it.

The right capital for buying another farm isn’t necessarily the right capital for transforming the one you already own.

Productivity is also wealth creation

If a farmer spends $500,000 and sustainably adds $100,000 to annual earnings, the benefit isn’t limited to the additional $100,000 of cashflow. Provided those earnings are sustainable and transferable, the enterprise itself has greater earning capacity and should ultimately be worth more as a result.

That value may not be fully captured in the land valuation, but it is still economic value. This is a different form of wealth creation from buying land and benefiting from an increase in the value of the underlying asset, but both can create equity.

Nor are the two strategies mutually exclusive. A more productive farm generates more cash, supports more debt and builds equity faster, which may ultimately provide the capacity to buy the neighbour anyway.

The next productivity cycle

New Zealand farmers have been very good at accumulating land, but we’ve also been remarkably good at making the land, animals and crops we already own more productive. The dairy and sheep numbers over the past 30 years demonstrate just how significant that contribution has been.

There is every reason to think there is considerably more to come across all three areas: better genetics and biological performance, better physical farming systems, and increasingly precise management using technology, automation and data.

And increasingly these opportunities will overlap.

Virtual fencing isn’t simply replacing a fence; it can change how much pasture we harvest. Genetics isn’t simply producing a better animal; it can change the labour, feed and cost structure of the farming system. AI and precision technology aren’t simply new pieces of machinery; they can allow us to manage individual plants, animals and square metres in ways that previously weren’t economic.

The common thread is that each has the potential to increase the earning capacity of the assets we already own.

Land expansion will remain part of the story, as will aggregation and conversion where the economics support them. But perhaps the starting question should become broader than simply: “What else can we buy?” A better question might be: “Where does our next dollar of capital create the most sustainable earnings?”

And for those of us involved in agricultural finance, there is another question alongside it: “Are we providing the right type of capital to fund it?”

Sometimes the best answer will still be over the neighbour’s fence. Increasingly, it may be on our side of it.