Platform

How We Calculate Equipment Values

March 1, 2026 7 min read IronValue Team

Transparency matters in any valuation system. If you're going to use a number to price a $200,000 trade-in, you should understand exactly how that number was derived. This post is a complete explanation of how IronValue generates valuation estimates - no black box.

Step 1: Data Sources

Every IronValue valuation starts with completed auction data. We ingest daily results from three major sources:

  • Ritchie Bros. Auctioneers - The world's largest industrial auctioneer publishes completed sale prices after each event. We collect these results for all heavy construction and earthmoving equipment categories.
  • IronPlanet - Online auction platform with guaranteed equipment inspections. IronPlanet's standardized condition reporting makes their data particularly reliable for cross-source comparisons.
  • Purple Wave - Regional auction leader with strong coverage of the Midwest and Plains states. Critical for dealers operating in those markets where regional pricing can differ from coastal averages.

We currently maintain over 10,000 completed auction records, updated daily. Our database goes back 24 months - far enough to capture market cycles without including stale data.

Step 2: Model Taxonomy and Variant Matching

This is where most competitor tools fall short. Raw auction data comes in messy: "CAT 320 GC", "Cat 320GC", "Caterpillar 320 GC Excavator" - all the same machine. We normalize and deduplicate this data against a structured equipment taxonomy.

More importantly, we track sub-model variants separately. A CAT 320 GC is a fundamentally different machine than a CAT 320 standard - different engine rating, different standard features, different market value. When you enter "320 GC," we search for 320 GC comps. Not just "320 family."

A CAT 320 GC is a different machine than a CAT 320. Comparable means actually comparable.

Step 3: Comp Filtering

Once we've identified the right make/model/variant, we filter the auction database for relevant comps:

  • Date filter: Last 24 months only. Older data introduces too much pricing cycle noise.
  • Hour range filter: ±40% of your machine's hours. For a machine at 4,200 hours, we look at comps from 2,520 to 5,880 hours. This prevents outliers from distorting the valuation.

After filtering, we typically have 10-50 candidate comps for common models like the CAT 320 GC or Deere 310SL. For rarer equipment, we may have fewer - which is reflected directly in the confidence score.

Step 4: Similarity Scoring

Each comp is scored for similarity to your specific machine across five weighted factors:

  • Hour meter proximity (40%): The closer the comp's hours are to your machine's hours, the higher it scores. This is the largest single factor.
  • Sale recency (25%): A sale from last month is more relevant than one from 20 months ago. More recent comps score higher.
  • Model year match (15%): Same year scores 100%; each year of difference reduces the score by approximately 10% of this weight.
  • Condition match (10%): Excellent-to-Good match scores higher than Good-to-Fair. Condition mismatches reduce comp relevance.
  • Geographic proximity (10%): Same state = full score. Same country (US or Canada) = half score. This reflects regional market variation.

Step 5: Weighted Valuation Calculation

We take the top comps by similarity score (typically 10-20) and calculate a weighted average of their sale prices, where each comp's contribution is proportional to its similarity score.

We also apply an hour-based adjustment: if the average hours of our top comps differ from your machine's hours, we apply a model-specific depreciation curve to adjust for those additional or fewer hours. This handles situations where available comps tend to be higher or lower hours than your specific machine.

Step 6: Price Range and Confidence

The estimated value is accompanied by a price range showing the 15th and 85th percentile of weighted comparable sale prices. This filters out statistical outliers while giving you a realistic spread of what the market will pay.

The confidence score reflects three factors:

  • Number of comparable sales found (more = better confidence)
  • Average similarity score of the top comps (higher similarity = better confidence)
  • Data freshness - how recently most of the comps sold

High confidence (≥85%) means we found 15+ comps with strong similarity scores and recent data. Low confidence means fewer comps, lower similarity, or older data - and the valuation should be treated with more caution.

What We Don't Do

A few important caveats:

  • We do not adjust for individual machine condition beyond the four-tier system (Excellent/Good/Fair/Poor). Specific mechanical issues require an on-site inspection.
  • We do not account for unusual or customized configurations. If your machine has a rare attachment combination or non-standard modifications, the valuation may not fully reflect those factors.
  • IronValue valuations are market estimates, not certified appraisals. For insurance, lending, or legal purposes, use a certified appraiser.

Want to see the algorithm in action? Try a demo valuation with sample data, or start a free trial to run valuations with live auction data.