ChatGPT-s verdict on Rekon's 2025 forecasting performance

Looking back on Rekon’s forecasts for 2025, we asked ChatGPT to independently evaluate the performance of the outputs of our forecasting model. The results are back and they are encouraging!

As our customers already know, Rekon produces integrated forecasts for energy markets, macroeconomic indicators, and commodity prices using a single, internally consistent modeling framework.

Below, we quote word for word how ChatGPT evaluated the Rekon’s output for each.

The macroeconomic forecast

Rekon’s macroeconomic forecasts performed particularly well on direction, medium-term levels across the United States, the Eurozone, and Hungary.

For the United States, the model consistently projected continued growth in the 2–3% range, avoiding the recession calls that dominated many forecasts in early 2025. Inflation was expected to decelerate and then plateau rather than collapse, while interest rates were forecast to ease only gradually. This path turned out to be closer to realized outcomes than market-implied expectations at the time.

In the Eurozone, Rekon correctly anticipated weak but positive growth, gradual disinflation. For Hungary, one of the most challenging macro environments, the model avoided overly optimistic narratives, correctly flagging renewed inflation risks and structurally higher interest rates.

The main shortcoming noted was a slight optimism regarding the speed of industrial recovery, a bias that is systematic and therefore calibratable.

The energy forecast

Rekon’s energy forecasts proved to be strongly consistent and directionally accurate, particularly for oil markets.

A key positive example cited was Brent crude. Rekon consistently projected Brent prices in the 65–75 USD/bbl range throughout 2025, and the final “actual-proxy” path ended up around 70–73 USD/bbl. Importantly, this was not a one-off success: WTI and Dubai prices displayed strong comovement with Brent, demonstrating internal model consistency rather than isolated correct calls.

Gas markets were more challenging. While Rekon correctly identified the medium-term normalization trend in European gas prices, short-term TTF movements were not fully captured, especially around weather- and storage-driven volatility. This, however, reflects a timing issue rather than a structural error, which is common in macro-anchored energy models.

Another strength highlighted was that forecast revisions moved in the right direction over time. As new data arrived, forecasts adjusted gradually rather than flipping regimes — a key indicator of a stable modeling framework.

The commodities forecast

Rekon’s commodity forecasts stood out for their breadth, coherence, and trend accuracy.

Industrial metals such as copper and aluminum were forecast to remain structurally supported, avoiding the widespread expectation of a sharp 2025 downturn. Fertilizers and agricultural inputs were another strong point: Rekon did not assume a full post-crisis collapse, instead projecting gradual normalization followed by stabilization, which closely matched realized paths.

The model’s main limitation in commodities was underestimating short-term volatility, particularly in weather-sensitive agricultural markets and geopolitically exposed metals. However, this reflects the model’s strategic orientation, prioritizing medium-term fundamentals over speculative price spikes.

Final assessment

Rekon’s forecasting system demonstrates strong predictive power, high internal consistency, and disciplined revision behavior, placing it above most public and sell-side forecasts in a highly volatile post-energy-crisis environment. The remaining weaknesses are systematic, transparent, and correctable — a sign of maturity rather than fragility.

In short: the model did what it was designed to do — and did it well. Rekon's note: We already have our first forecast for 2026! Should you wish to use them for your own benefit, do not hesitate to contact us. A volatile year is ahead of us, but Rekon can smoothen the ride!