Collect respondent prices
Respondents submit fair EOD price assessments for the relevant commodity, basis and delivery period.
Methodology
How SPIKE SPOT INDEX is calculated
SPIKE SPOT INDEX uses a repeatable daily process that turns respondent-partner and Monitor assessments into verified and representative index values for each position, basis and calculation date.
Calculation approach
From partner and Monitor assessments to a published spot benchmark
The index is calculated from daily price assessments submitted by a defined group of Spike Brokers respondent partners. Each submitted value should reflect a fair end-of-day market level for the relevant commodity and basis.
MN7R Monitor is connected as a separate automated respondent. Every weekday at 17:00 Kyiv, SPIKE imports current Monitor price indicators and matches them by commodity, delivery basis and delivery-period overlap. A Monitor record is eligible only when its delivery period overlaps the current 30-day window by at least 10 days.
If several Monitor indicators match one SPIKE position, the system calculates their arithmetic average and stores it as the Monitor respondent value for that position. The public structure is split into Grains Export, Oilseeds crush and Oilseeds Export. Core and seasonal export positions use CPT Port or FCA Chop bases, while processing positions use CPT Crush.
Prices are validated by sample median. Values deviating from the median by more than +/-2% are excluded as potential outliers. The final index value is the arithmetic average of the cleaned sample. A basket is publishable only when at least 5 valid respondent prices remain after filtering.
Before publication, values can be reviewed, corrected and versioned. After publication, the final value is locked. The system records changes, recalculations and publication events in an audit log. Public methodology identifies respondent types, but not company names.
CPT Port
Core and seasonal export basis
FCA Chop
Border-position basis
CPT Crush
Processing-position basis
3
Index categories
17:00
Monitor import time (Kyiv)
10d / 30d
Minimum delivery-window overlap
5+
Minimum valid respondents
+/-2%
Median-based outlier filter
Locked
Final value after publication
Methodology
Calculation flow
Validate the sample
The system calculates the median and checks all submitted values against the central market level.
Exclude outliers
Prices deviating by more than ±2% from the median are excluded from the cleaned sample.
Calculate the index
The published value is the arithmetic average of the cleaned sample, provided the minimum respondent count is met.
Verify and lock
Before publication, values may be reviewed and versioned. Once published, the final value is locked and recorded in the audit log.
AI usage policy
How AI is used in SPIKE SPOT INDEX
SPIKE SPOT INDEX uses a hybrid approach: deterministic index calculation first, AI-assisted interpretation second.
The official index calculation does not rely on AI. Published values are calculated through the SPIKE methodology: respondent data collection, median validation, outlier filtering, arithmetic averaging of the cleaned sample, minimum respondent coverage and locked publication.
AI is used only as an analytical layer above published or preview data. The analytics page explains index movement, volatility and spreads; Context separately monitors external sources and generates daily / weekly / monthly market context.
AI outputs and Context reports are not official prices, trading signals or recommendations. They do not access individual respondent submissions and do not override the methodology. All AI-assisted content should be read as analytical context for market understanding.
01
verified market data first
02
deterministic methodology second
03
Context market context third
Official SPIKE SPOT INDEX values are methodology-based and non-AI-generated. AI-assisted outputs are provided for analytical context only.
Operational reliability
How publication quality is controlled
The platform separates official calculation, publication, analytical interpretation and external delivery channels. This reduces accidental changes to locked values and makes the operational state verifiable before release.
01
Methodology controls
Official values pass sample validation, outlier filtering and post-publication locking.
02
Change log
Recalculations, manual corrections, catch-up runs and publication events are kept as an audit trail.
03
Release gates
Before release, the code passes repo audit, production-env smoke, lint, test suite and production build.
04
Tenant boundaries
Public sites, APIs, Context and messaging channels use separate tenant/env boundary checks.
Production secrets, individual respondent submissions and internal operational keys are not published.
Official methodology PDF
Download the condensed SPIKE SPOT INDEX methodology in English. The Ukrainian version is available on the UK page.
Methodology
FAQ
What is SPIKE SPOT INDEX?+
It is a daily Spike Brokers spot benchmark for selected Ukrainian agricultural export and processing positions.
Who provides data?+
Data are collected from a defined pool of Spike Brokers respondent partners and from MN7R Monitor as a separate automated respondent. Individual company submissions are not disclosed in public outputs.
How is MN7R Monitor used in the calculation?+
Each weekday at 17:00 Kyiv the platform imports current Monitor indicators. Matching is based on commodity, delivery basis and delivery-period overlap of at least 10 days inside the active 30-day window. If several indicators match one SPIKE position, their arithmetic average becomes the Monitor respondent value.
How are outliers handled?+
The system calculates the median of the respondent sample and excludes values that deviate by more than ±2% from that median before calculating the final average.
What happens if there is insufficient data?+
If fewer than 5 valid respondent prices remain after filtering, the basket is not publishable as an official index value.
Are respondent companies published in the methodology?+
No. The public methodology describes respondent types such as brokers, traders, exporters and processors, but does not disclose the names of individual companies.
Can published values be changed?+
Before publication, values can be reviewed and versioned. After publication, the final value is locked and changes are recorded through the audit trail.
Can the methodology scale?+
Yes. The methodology supports future weighted baskets, additional delivery periods such as T+60 and T+90, and separate bid, offer or mid quote types.
Does AI calculate the official SPIKE SPOT INDEX value?+
No. Official index values are calculated according to the published SPIKE methodology. AI is used only to interpret published data, generate analytical briefs, describe volatility and produce scenario notes. It does not set, adjust or validate official index values.