The price-volume split in Azerbaijan
Jan to Jul 2026 customs data already shows the core signal buyers need to price in: Azerbaijan exported about 8,900 tons of hazelnuts worth $100.2 million, with export volume down 22.5% year on year while export value rose by about 22%. When volume falls and value rises, the export unit value is moving up fast, but that does not automatically mean farmgate prices are up by the same amount.
A higher average $/t can be a mixed grade effect. If more kernel vs in-shell is shipped, or if exporters prioritize industrial kernels with better calibers and tighter specs, the unit value rises even if farmgate is choppy. The same can happen with shrinkflation in specs, where the contract still says “standard” but the lot is effectively a cleaner, more sorted selection because only those lots can pass buyer requirements and border controls.
Selective exporting also changes what EU buyers see. If exporters ration supply and focus on the best performing lots, the market can look bullish in offers even while some districts report softer harvest-week prices. That is why producer price snapshots and local anecdotes should not be your main benchmark. Delivered kernel offers into the EU (DDP or at least CIF plus realistic clearance costs) are the numbers that matter for replacement cost.
Do the buyer math early, not after a line stop. If export unit values are rising while contract volumes are cut, the risk is not only higher €/kg, it is physical availability. Model “tons at risk” per month against safety stock. A confectionery plant consuming 80 to 120 tons/month can lose negotiating power quickly if even one container slips and there is no qualified alternative grade.
Answer three questions before you call the market bullish. Is the exporter mix changing, with only a smaller set able to ship compliant lots? Are we seeing supply rationing where better calibers go to the EU and lower lots move elsewhere? Are offers quietly shifting format, for example more kernel and less in-shell, which lifts the headline unit value? Procurement actions follow from that: split tenders by grade and format, and insist every offer states crop-year plus a defect schedule, not just a price.
Use one KPI checklist and keep it current, because divergence across KPIs often shows up before disputes do:
- Export volume trend
- Export unit value trend
- RASFF border outcomes
- Lead times
Supply tightness drivers to test
“Tight supply” headlines are not enough to buy on. Buyers need to validate orchard yield variability, biennial bearing, labor availability for harvest, and post-harvest drying and storage capacity. In hazelnuts, moisture management is not a side issue. It is directly linked to mold and mycotoxin risk, and it is often where stressed supply chains break.
Recent research highlights why you should ask weather questions, not just price questions. A 2026 paper in Frontiers in Plant Science discusses pre-harvest dynamics of Aspergillus section Flavi and links risk to climate variables such as temperature, rainfall, and humidity windows. Translate that into practical due diligence: what was the weather pattern in key regions during kernel fill and in the pre-harvest period, and did it force delayed pickup or slow drying?
Loss points are usually operational, not agricultural. Slow drying after pickup, mixed lots from multiple smallholders, and prolonged storage in non-controlled warehouses are common pathways to hidden mold and later aflatoxin findings. “Good” looks like disciplined intake moisture checks, documented drying time, aeration, pest control, and FIFO with traceable lot movement. If a supplier cannot show records, you should assume variability, even if the sample looks fine.
Verification should also include anti “paper supply” controls. Ask for yield and area evidence such as grower registers, orchard GPS blocks, and purchase receipts, then reconcile that with mass balance. In-shell to kernel conversion rates should be plausible and consistent with the supplier’s process. When export tonnage is down, overselling risk rises, and mass balance is one of the few tools a buyer can use before the shipment fails.
Quality pressure tends to show up as a cluster, not a single defect. Brown rot, kernel rancidity, hidden mold, higher defect rate, foreign matter, and insect damage often increase together when harvest and drying are rushed. In a stressed market, tighten defect tolerances and be explicit about how defects are measured, because “standard” can mean different things across exporters.
The decision impact is simple. Supply tightness plus inconsistent post-harvest controls increases the probability of spec nonconformance. Build a quality fallback plan now, with dual approved suppliers and an alternative grade approved for specific SKUs, so you are not forced into emergency rework.
Export reality check on EU grades
The export contraction suggests availability is becoming selective. Jan to Jun 2026 exports were about 8,341 tons, down 22.4% year on year by volume, while export value rose 24.6%. In practice, that often means EU buyers see cleaner lots and more capable exporters, while borderline lots are diverted to other channels or markets with different tolerances.
EU industrial terminology is also where misunderstandings start. Buyers usually specify raw kernels, blanched kernels, diced or granulated, paste, roasting grade, and chocolate grade. Azerbaijan can offer kernels and processed formats, but EU processors often hit constraints around consistent calibration, low defects, and homogeneous lots across repeated shipments. If you need uniformity for automated dosing or whole-kernel inclusions, you should treat calibration consistency as a supply constraint, not a nice-to-have.
Ask grade availability questions that force a real answer. Can the supplier supply 11 to 13 mm or 13 to 15 mm kernels consistently across multiple lots, not just one container? What are typical defect limits for mold, shrivel, rancid, and insect damage, and how are they tested? Do they segregate by region or collector to reduce cross-contamination, or is it bulk blending?
Inspection reality matters for planning. A quantitative EU briefing notes increased controls for aflatoxin on hazelnuts from the region, which affects lead times and costs through risk-based sampling. Even when a lot is compliant, border sampling can add time, and that time is part of your true landed cost.
For a confectionery buyer needing 3 to 5 containers/month of uniform kernels for praline, the operational play is straightforward. Require a pre-shipment COA per lot plus retention samples, and plan for possible border sampling delays when risk flags are active. If your production plan assumes just-in-time arrivals, you are effectively betting against the inspection environment.
Structure RFQs to reflect how the market actually clears. Separate lines for “direct consumption compliant lots” versus “for further processing (sorting) lots,” with distinct aflatoxin limits and clear remediation options. That makes it easier to accept a usable lot without forcing it into a spec it cannot meet.
Verification before switching origin
Switching origin is mostly a traceability and compliance project, not a pricing project. Request a farm or collector list, lot coding from intake through cracking, sorting, and packing, plus mass balance and evidence of segregation by crop year. Ask whether the chain of custody is identity preserved or bulk, and confirm the one-step-back one-step-forward records are complete.
Residue compliance needs to be explicit. Require a current pesticide residue program, EU-aligned multi-residue screening, an MRL compliance statement, and clarity on lab scope and testing frequency. Confirm the lab is accredited to ISO 17025 and that methods cover relevant actives used in hazelnut orchards, because “we test residues” is not the same as “we test what the EU will enforce.”
Aflatoxin control is the gating risk. EU maximum limits are set under Regulation (EU) 2023/915, and the operational implication is that limits depend on intended use, such as direct consumption versus sorting or further processing. Contract to the correct limit for your use case, and do not accept vague language like “EU compliant” without the numeric target and method.
Enforcement data supports tougher incoming controls. RASFF records show multiple border rejections for Azerbaijan hazelnuts in recent quarters, often for aflatoxin B1 or total aflatoxins. That increases the probability of delays, rejections, and extra sampling costs, even for buyers who have never had an issue before.
Your playbook should be non-negotiable. Mandate pre-shipment COA, define a sampling protocol that uses incremental samples across bags and a composite per lot, keep retention samples, and set dispute resolution in advance, including which lab result prevails. If you leave the lab hierarchy undefined, you will argue about it when the container is already blocked.
Supplier audits should focus on controls that prevent problems, not paperwork. Review drying parameters, storage RH and temperature monitoring, sorting technology including optical sorting and any aflatoxin-focused sorting steps, HACCP, corrective actions from prior nonconformities, and documented release criteria.
Contracting implications for 2026-27
Price formulas should reflect grade mix, because value up and volume down can hide what is actually being shipped. Use spec-linked matrices that price by caliber, defects, and processing format, rather than a single headline €/kg. If you buy in-shell, include kernel ratio and yield-adjusted pricing, because the conversion is where cost surprises appear.
Delivery schedules need buffers. Increased border controls and higher mycotoxin scrutiny imply longer and less predictable clearance. Build schedules with shipping windows plus inspection contingencies, often adding 2 to 4 weeks, and define who pays demurrage if held for official sampling. If you do not allocate that risk, it will be allocated to you by default.
Default risk rises when export tonnage is lower. Suppliers may oversell, or they may prioritize higher-priced buyers. Mitigate with staged payments, performance bonds where feasible, and force majeure language that distinguishes crop failure from commercial nonperformance.
Quality and food safety clauses should be written for enforcement, not goodwill. Specify EU limit compliance, define rejection rights, and require notification of any RASFF-related events connected to the exporter or upstream aggregators during the contract period. A practical clause set is: lot is released only upon pre-shipment COA, arrival inspection passing aflatoxin and defect limits, and complete traceability pack list plus lot codes. If arrival fails, require replacement within an agreed timeframe or a credit note at the buyer’s option.
Risk-sharing for 2026 to 2027 usually means splitting origins and splitting delivery months. Consider call-off contracts where volumes are optional within a band, such as plus or minus 20%, to reduce exposure to sudden export disruptions.
Azerbaijan versus Turkey and Georgia
Turkey remains the dominant benchmark origin for Black Sea supply because of scale, exporter depth, and broad grade availability. Azerbaijan and Georgia tend to be supplemental sources for specific specs, timing, or diversification. An EU quantitative briefing puts EU-27 hazelnut import value shares in 2023 at about 70% Turkey, about 6% Georgia, and about 5% Azerbaijan, which is a useful reminder that Azerbaijan typically cannot replace Turkey at scale for large industrial programs.
Use a simple decision rule. Azerbaijan is a good tool for diversification, spot gaps, and certain kernel specs when offered competitively. It is a risky choice as a sole origin if you require uninterrupted multi-container monthly flow.
Quality risk comparisons should be framed carefully. Increased aflatoxin attention affects the region, so the differentiator is process control maturity, not the country label. Qualify supplier by supplier, focusing on drying discipline, segregation, and testing regime.
Blending strategies can reduce exposure without lowering standards. Many processors qualify Azerbaijan lots as part of a blend spec for paste or praline where minor sensory variance is acceptable, while keeping strict single-origin requirements for premium whole-kernel applications.
Procurement planning should be built as a three-origin matrix. Decide minimum monthly guaranteed tons plus optional tons for Turkey, Georgia, and Azerbaijan, then map which origin gives the best security for Q1 to Q2, where border hold risks are highest, and which suppliers can cover calibers when Turkey is tight.