Skip to content
Alternative Finance

Underwriting · Direct funders

Loan portfolio concentration calculator

Paste one exposure per line with its balance, industry and state and this loan portfolio concentration calculator returns the Herfindahl index across borrowers, the largest single exposure and top five as a share of the book, the largest industry and state, a full industry table, a state chart and a list of every breach of your single-exposure, industry and state limits. It is for direct funders managing a book and for the lenders and investors who finance them. Watch the industry table: a book that looks diversified by borrower count is often two or three trades in disguise.

  • Largest exposure is 19.3% of the book, above the 10% limit.
  • Construction is 31.9% of the book, above the 25% industry limit.
Herfindahl index (borrowers)
1,259
Largest single exposure
19.3%
Top 5 exposures
72.7%
Largest industry
Construction: 31.9%
Largest state
FL: 24.8%
Total book
$2,180,000
Exposures
10

The Herfindahl index sums squared shares (×10,000); below 1,000 is diversified, above 1,800 concentrated. A book of equal exposures to N borrowers scores 10,000 ÷ N.

Share by state
  • FL24.8%
  • TX20.9%
  • NY17.9%
  • OH15.1%
  • GA14.2%
  • CA4.4%
  • NJ2.8%
Industry concentration
IndustryBalanceShare
Construction$695,00031.9%
Trucking$490,00022.5%
Restaurants$370,00017.0%
Manufacturing$330,00015.1%
Medical$140,0006.4%
Retail$95,0004.4%
Salons$60,0002.8%

Formula

Total book = sum of all balances. Share of each exposure, industry or state = its balance ÷ Total book.

Herfindahl index (borrowers) = 10,000 × Σ (exposure share)² across all exposures; a book of N equal exposures scores 10,000 ÷ N.

Largest single exposure = largest balance ÷ Total book; Top 5 exposures = sum of the five largest ÷ Total book.

Breach flags: largest exposure above Policy max single exposure; any industry above Policy max per industry; any state above Policy max per state.

Worked examples

Example 1: the default ten-exposure book

Inputs: ten lines totalling $2,180,000: Restaurants NY $250,000; Trucking TX $180,000; Construction FL $420,000; Retail CA $95,000; Trucking GA $310,000; Medical NY $140,000; Salons NJ $60,000; Construction TX $275,000; Restaurants FL $120,000; Manufacturing OH $330,000. Policy limits: 10% single, 25% industry, 30% state. Largest single exposure = 420,000 ÷ 2,180,000 = 19.3%. Top 5 = 420,000 + 330,000 + 310,000 + 275,000 + 250,000 = 1,585,000 ÷ 2,180,000 = 72.7%. Industry: Construction 695,000 = 31.9%; Trucking 490,000 = 22.5%; Restaurants 370,000 = 17.0%; Manufacturing 330,000 = 15.1%; Medical 140,000 = 6.4%; Retail 95,000 = 4.4%; Salons 60,000 = 2.8%. State: FL 540,000 = 24.8%; TX 455,000 = 20.9%; NY 390,000 = 17.9%; OH 15.1%; GA 14.2%; CA 4.4%; NJ 2.8%. Herfindahl: square each borrower share and sum: 0.1147² + 0.0826² + 0.1927² + 0.0436² + 0.1422² + 0.0642² + 0.0275² + 0.1261² + 0.0550² + 0.1514² = 0.1259, times 10,000 = 1,259. Two breaches: the largest exposure at 19.3 percent exceeds the 10 percent limit, and Construction at 31.9 percent exceeds the 25 percent industry limit. No state exceeds 30 percent, though Florida is close.

Example 2: five equal exposures

Inputs: five lines of $200,000 each, in Retail NY, Medical CA, Trucking TX, Restaurants FL and Manufacturing OH; default limits. Total book $1,000,000. Each exposure is 200,000 ÷ 1,000,000 = 20.0%, so the largest single exposure is 20.0% and the top five are 100.0%. Each industry and each state is also 20.0%. Herfindahl = 10,000 × 5 × 0.20² = 10,000 × 5 × 0.04 = 2,000, which matches the rule that N equal exposures score 10,000 ÷ N. One breach: the 20 percent single exposure exceeds the 10 percent limit. This book is perfectly spread across industries and states and still concentrated, because it has only five borrowers; the Herfindahl index of 2,000 is above the 1,500 level the tool flags. The lesson is that diversification is about count and size as much as category. To bring the HHI below 1,000 with equal exposures, the book needs more than ten borrowers; to bring the largest exposure under 10 percent, it needs at least ten.

Example 3: one dominant exposure

Inputs: Construction FL $900,000; Retail FL $100,000; Medical GA $100,000; Salons NY $100,000; default limits. Total $1,200,000. Largest single = 900,000 ÷ 1,200,000 = 75.0%. Top five = 100.0%. Industry: Construction 75.0%, the other three 8.3% each. State: FL 1,000,000 = 83.3%; GA 8.3%; NY 8.3%. Herfindahl = 10,000 × (0.75² + 3 × 0.0833²) = 10,000 × (0.5625 + 0.0208) = 5,833. Three breaches: the largest exposure at 75 percent against a 10 percent limit, Construction at 75 percent against 25 percent, and Florida at 83.3 percent against 30 percent. This is what a book looks like when a funder makes one large deal to a relationship borrower: the book's loss experience is no longer a portfolio statistic but a single credit decision, and a Florida construction downturn would impair 83 percent of it in one event. The remedy is not more small deals around the edges but a participation or sale of part of the large exposure.

Why concentration is a separate risk from credit quality

Expected loss is an average: a 7 percent charge-off rate means that over a large, diverse book about 7 percent of funded volume will be lost in a typical year. The average only arrives smoothly if the losses are independent. When many exposures share a driver, such as the same industry, the same state or the same large borrower, losses arrive together, and a book that was adequately reserved for its average can be insolvent after one correlated event. A loan portfolio concentration calculator measures that correlation risk by asking how much of the book depends on any one borrower, trade or geography. Bank examiners treat concentrations of credit as a distinct supervisory topic, covered in the OCC's Comptroller's Handbook and the FDIC's examination guidance, and warehouse lenders write concentration limits into every facility.

The three dimensions in the tool are the ones that matter most for small-business lending. Single-borrower concentration is the risk that one default is material. Industry concentration is the risk that a trade-wide shock, fuel prices for trucking, a materials spike for construction, a lockdown for restaurants, hits many borrowers at once. State concentration captures regional economic shocks, natural disasters, and, increasingly, state regulatory change: a book heavy in California or New York is exposed to those states' commercial financing disclosure regimes in a way a spread book is not. The state pages summarise the regulatory environment where a funder's exposure is concentrated.

Reading the Herfindahl index
HHIEquivalent equal borrowersReading
Under 1,000More than 10Diversified by borrower
1,000 to 1,8006 to 10Moderately concentrated; the tool flags above 1,500
Above 1,800Fewer than 6Concentrated; one or two defaults are material
Above 5,000Fewer than 2A single credit, not a portfolio

The Herfindahl index and what it adds to the top-borrower share

The Herfindahl-Hirschman index was developed to measure market concentration and is used by the U.S. antitrust agencies, whose explanation of the index is the standard reference; the Wikipedia article gives the mathematics. Applied to a loan book, it sums the squared shares of every exposure, so large exposures count far more than small ones: a 20 percent exposure contributes 400 points, a 2 percent exposure contributes 4. The result equals 10,000 divided by the number of equal-sized borrowers that would produce the same concentration, which makes it intuitive: a book with an HHI of 1,259, as in the default example, is as concentrated as a book of about eight equal borrowers, even though it has ten.

The largest-exposure and top-five shares describe the tail; the HHI describes the whole distribution. A book can pass a 10 percent single-exposure limit and still have an HHI above 1,500 if it has only a dozen borrowers of similar size, and a book can have an HHI below 1,000 while one exposure breaches the limit if it has hundreds of small loans and one large one. Use the single-exposure limit to control the tail and the HHI to track overall granularity as the book grows. The tool computes the HHI across borrowers only; compute industry and state HHIs by hand from the table and chart if your policy requires them.

Setting policy limits

The three limits are policy inputs because the right numbers depend on the size and purpose of the book. For a new funder with a small book, a 10 percent single-exposure limit is often impossible to meet and a 20 to 25 percent limit with a path to 10 percent as the book grows is realistic. Industry limits of 20 to 30 percent and state limits of 25 to 40 percent are common in warehouse facilities for alternative lenders, with lower limits for industries the lender considers volatile. Many facilities also impose limits on the top ten exposures combined, on balances above a stated size, and on exposures to renewals of the same merchant, which this tool does not compute but which can be read from the sorted list.

  • Set single-exposure limits as a share of equity as well as of the book, since equity is what absorbs the loss; a 10 percent exposure in a book funded 80 percent by debt is 50 percent of equity.
  • Set tighter industry limits for trades with high loss severity in your own data, and revisit them when the loss experience changes.
  • Treat the same borrower under different entity names, and affiliated borrowers under common ownership, as one exposure.
  • Decide in advance what happens on a breach: no new funding in that category until the share falls, participation of the excess, or a documented exception with an expiry date.

The Census Bureau's Annual Business Survey describes the distribution of U.S. businesses by industry and state, which is a reasonable reference for what a representative book would look like; a funder whose book is 32 percent construction when construction is a far smaller share of the small-business population has chosen that concentration, deliberately or through its broker channel, and should price and reserve for it.

Preparing the data and reading the outputs

Each line is one exposure: the current balance, the industry label and the two-letter state, separated by commas. Use the outstanding balance or remaining payback, not the original amount, so that the picture reflects current exposure; the loan data CSV validator checks an export for the formatting problems that break the paste. Industry labels should come from a fixed list so that the same trade is not split across spellings. For funders with several hundred exposures, paste the full list; the industry table and state chart aggregate, and the top-five and HHI figures are only meaningful on the whole book.

The outputs feed three decisions. The breach list is an immediate action item for the credit committee and the warehouse lender's compliance certificate. The industry table shapes origination targeting: a funder at 32 percent construction should instruct its brokers accordingly or price construction higher. The HHI and top-five shares set the diversification goal and the rate at which the book should add new borrowers rather than renewing existing ones. Combine the result with the loan loss reserve calculator, adding a qualitative overlay for concentrated segments, and with the loan delinquency rate calculator, run by industry, to see whether the concentrated segment is already underperforming.

Reducing concentration

Concentration falls in only two ways: the large exposures shrink or the rest of the book grows. Amortization shrinks large exposures naturally in a short-term book, which is why a breach caused by one large deal cures itself within the term if no renewal is written. Selling or participating a share of a large exposure to another funder is the standard remedy for a breach that cannot wait; the participating funder takes a pro-rata share of the balance and the collections, and the originating funder keeps the relationship. Growing the rest of the book is the slow cure and the one that works for industry and state concentration, by directing broker and marketing effort toward under-represented trades and regions.

The failure mode to avoid is diluting concentration by lowering underwriting standards in the under-represented categories; a book that adds weak retail exposures to offset strong construction ones has traded concentration risk for credit risk. Use the merchant risk score calculator and the same policy thresholds in every category, and accept that a diversified book is built over several turns. The loan portfolio yield calculator shows the cost of carrying a participation or of slowing renewals, and the Federal Reserve's charge-off release illustrates, in the bank data, how loss rates by loan category diverge in stress periods, which is the scenario concentration limits exist to survive.

Questions people ask

What is a good Herfindahl index for a loan portfolio?

Below 1,000 indicates a book as diversified as more than ten equal borrowers; the tool flags values above 1,500, and above 1,800 is concentrated by the convention used in antitrust analysis. A book of N equal exposures scores exactly 10,000 ÷ N, so a twenty-borrower book of equal size scores 500. Large single exposures raise the index sharply because shares are squared.

How much of a portfolio should one borrower represent?

Common policy limits run from 5 to 10 percent of the book for established funders, with higher interim limits for small or new books. Because losses are absorbed by equity, also express the limit as a share of equity: a 10 percent exposure in a book funded 80 percent by debt equals half the funder's equity, which most investors would consider too much for one credit.

Why does the loan portfolio concentration calculator flag industry and state separately?

Because they represent different correlated shocks. An industry shock such as a fuel price spike hits trucking borrowers in every state; a regional shock such as a hurricane or a state regulatory change hits every industry in one state. A book can be well spread on one dimension and dangerously concentrated on the other, so each needs its own limit and its own breach list.

Should I enter original amounts or current balances?

Current balances or remaining payback, because concentration risk is about what you could lose today. A large deal written six months ago on a nine-month term is already a third of its original size. Using current figures also shows how quickly a breach will cure through amortization if no renewal is written, which often makes a participation unnecessary.

How do I handle several advances to the same merchant?

Combine them on one line before pasting, and combine affiliated entities under common ownership the same way. The single-exposure test is meant to capture the loss from one borrower's failure, and a merchant with three positions totalling 15 percent of the book is a 15 percent exposure regardless of how the contracts are papered.

Sources and further reading

Related tools and guides

More underwriting

Updated October 12, 2026 · By Alternative Finance editorial team

Next step

Turn the estimate into real offers

One request, several funding partners, side-by-side payback. Nothing is binding until you accept an offer.

  • Soft inquiry to request offers; a hard pull only if you proceed with a funder.
  • Decisions in as little as 24 hours for many products; funding often within days.
  • Use our calculators to check the true cost before you accept anything.

See what you may qualify for

Takes about a minute. Requesting offers does not affect your credit score. Free, no obligation.

We are not a lender or broker. Requests are forwarded to funding partners free of charge. How the site is paid.