What makes a useful
soybean analog?

The question comes first.

A balance sheet is a set of connected questions: how much supply is available, where demand is changing, and what remains in inventory. An analog can help organize those questions by comparing one set of changes with other sets. Its usefulness depends on the comparison being made explicit.

Forest View’s explorer focuses on the transition between crop years. It compares changes in beginning stocks, production, domestic crush, exports, and ending stocks, with the prior-year change in exports providing additional context. A match on one variable alone can obscure a very different overall balance.

A worked scenario.

Consider a purely illustrative case in which production rises 3%, crush increases 4%, exports fall 6%, and ending stocks decline 12%. Beginning stocks and prior-year export growth are held at +2% and −3%. These are independent demonstration inputs; the explorer does not enforce a physical mass balance.

The model ranks 20 synthetic comparison scenarios using normalized differences across the six input features. The five closest scenarios form the baseline comparison set. A smaller distance indicates a closer input match. It does not indicate the probability that a scenario will repeat.

The median path change in this demonstration is +1.7%. The middle half of endpoint outcomes lies between +1.4% and +1.7%. Because the paths are generated demonstration sequences, these figures provide no evidence about actual returns, forecasting skill, or expected future prices.

Read the dispersion, not just the line.

A single median can conceal disagreement among the comparisons. The explorer retains individual paths and displays the middle 50% of the selected set. That shaded area is a descriptive range within a small sample, not a confidence interval.

Change export growth or ending stocks and rerun the model. Observe which scenarios move into the comparison set and which leave it. Sensitivity to a small assumption change is itself useful context: it can identify where an analytical conclusion is fragile.

What the model cannot tell you.

• Similar balance-sheet changes do not establish that the same economic mechanism is at work.
• Policy, weather, timing, price levels, and market structure can differ even when a balance-sheet comparison is close.
• This demonstration does not account for revisions, point-in-time data availability, transaction costs, or out-of-sample validation.
• No actual market data or historical-year price path is included. No performance claim can be drawn from these results.