HHI Calculator (Herfindahl-Hirschman Index)

HHI is the sum of squared market shares (in percent points). Square each firm’s market share and add them up. Maximum is 10,000 (one firm with 100% share). Minimum approaches 0 (infinitely many firms with infinitesimal share).

HHI = Σ(s_i)²

Where s_i is firm i’s market share as a percentage (e.g., 30 for 30%).

Why squaring matters. A market with two firms at 50% each has HHI = 2,500 + 2,500 = 5,000. A market with ten firms at 10% each has HHI = 10 × 100 = 1,000. The squaring punishes concentration heavily, which is the whole point. Large firms dominate market dynamics far more than their share alone suggests.

DOJ / FTC merger guidelines (2023):

• HHI < 1,000: unconcentrated. Mergers here rarely raise structural concerns.
• HHI 1,000-1,800: moderately concentrated. No structural presumption, but a large increase still invites questions.
• HHI above 1,800: highly concentrated. A merger that also raises HHI by more than 100 points is presumed illegal.

Note the two conditions work together. The presumption needs both a post-merger HHI above 1,800 and a change above 100, not either one alone. There is a second route to the same presumption that has nothing to do with HHI: a merged firm holding more than a 30% share combined with a 100-point increase.

The 2023 guidelines tightened the old 2010 thresholds, which set the highly-concentrated cutoff at 2,500 and the moderately-concentrated floor at 1,500. If you see 2,500 quoted as the line, that is the superseded standard. Both agencies have also signalled they will challenge deals below these numbers where the evidence supports it, so treat the thresholds as a screen rather than a safe harbour.

Typical real-world HHI values.

• US wireless carriers (Verizon, AT&T, T-Mobile): ~3,000+ (highly concentrated)
• US auto manufacturers: ~1,600 (moderately concentrated)
• US grocery (national): ~750 (unconcentrated nationally, but local markets often >2,500)
• US airlines: ~1,400 (moderately concentrated)
• US oil refining: ~750
• US software (broadly): ~250 (very fragmented across categories)

The “national vs local” trap. Antitrust authorities care about the relevant geographic market. A national HHI of 800 might mask local HHIs of 4,000+ in specific cities. The 2007 Whole Foods / Wild Oats merger was challenged on local market HHI even though national HHI was modest.

Worked example. Six firms in a market:

• Firm A: 35%
• Firm B: 25%
• Firm C: 15%
• Firm D: 10%
• Firm E: 10%
• Firm F: 5%

HHI = 35² + 25² + 15² + 10² + 10² + 5² = 1,225 + 625 + 225 + 100 + 100 + 25 = 2,300

At 2,300 that market is highly concentrated under the 2023 guidelines, so any merger between two of the larger firms would draw serious antitrust review.

Pre-merger and post-merger HHI. When firm B (25%) merges with firm C (15%), the combined firm has 40% share. New HHI:

• New HHI = 40² + 35² + 10² + 10² + 5² = 1,600 + 1,225 + 100 + 100 + 25 = 3,050
• Change in HHI = 3,050 - 2,300 = +750

A 750-point increase in a market already at 2,300 clears both prongs of the structural presumption by a wide margin, and the combined firm’s 40% share clears the separate 30% route as well. That deal gets challenged.

Quick math shortcut. When two firms with shares s1 and s2 merge, ΔHHI = 2 × s1 × s2. For 25% and 15%: ΔHHI = 2 × 25 × 15 = 750. (Same answer as full recalculation.) This is how antitrust attorneys quickly screen deal HHI impact.

Limitations.

• HHI ignores entry barriers, product differentiation, switching costs, and pricing dynamics.
• A market can be technically “unconcentrated” by HHI but still have problematic dynamics (e.g., implicit collusion, captive customers).
• Defining the market is more art than science. Lawyers spend more time on market definition than HHI math.