Essay

Risk premium: what doubt costs, and what 107 years of bond data show

Published · Updated 17 min read

A risk premium is the extra return people demand for taking a risk, over what they could earn from a safe alternative. The best known version is the equity risk premium, which the NYU Stern professor Aswath Damodaran describes as the "extra return" investors collectively demand for putting their money into stocks instead of a riskless or close to riskless investment. It reflects both hopes and fears, "rising as the fear factor increases."[1]

This essay brings together three bodies of evidence. The first is what markets charge today for different levels of doubt: the extra interest a weaker borrower pays, and the extra return investors want in less familiar countries. The second is economic research on what happens when buyers cannot judge quality, and on the guarantees that fix it. The third is our own study of US bond yields and bank lending since 1919, which shows how the price of risk moves when fear rises. We close with a framework, the Doubt Ledger, and a method any supplier can use to see, shrink and charge for risk.

Idea in brief

The problem. Every buyer who considers a supplier it does not know yet charges a risk premium. It rarely says so. The premium shows up as a discount, a long decision, a small first order or slow payment.

Why it happens. When buyers cannot judge quality before they buy, they discount everyone, as George Akerlof showed in 1970.[5] Lenders do the same in plain numbers: in January 2026, a borrower rated B paid a default spread of 3.21%, eight times the 0.40% paid by a AAA borrower.[2]

The gap (our data). The price of doubt is not fixed. In every one of the 11 US recessions since 1953, the gap between the yield on Baa-rated corporate bonds and 10-year Treasuries widened, with a median rise of 85% from the month before the recession to the peak. Net bank tightening on business loans went from 0.0% in early 2007 to 83.6% in late 2008.

The answer. Shrink the premium buyers charge you with proof they can check, guarantees that move risk onto you, and a visible track record. Charge for the risk you carry with deposits, scope and payment terms. Expect both to swing when the economy turns.

Key findings

The essay in 3 minutes
  1. Finding 18.85%

    Doubt is priced in steep steps.

    In Damodaran's January 2026 table, the default spread was 0.40% for AAA, 1.11% for BBB, 3.21% for B and 8.85% for CCC.[2]

  2. Finding 212.64%

    New markets carry a premium.

    In January 2026, Damodaran's equity risk premium was 4.23% for Germany, 5.33% for Poland, 6.69% for Mexico, 7.47% for Brazil and 12.64% for Nigeria.[3]

  3. Finding 36.43%

    The premium jumps when fear rises.

    Damodaran's implied equity risk premium for the S&P 500 rose from 4.37% at the end of 2007 to 6.43% at the end of 2008.[4]

  4. Finding 485%

    Credit spreads widened in all 11 recessions since 1953.

    The Baa minus 10-year Treasury spread rose by a median of 85% from the month before each recession to its peak. In 2008 it went from 2.55 to 6.01 percentage points.[7]

  5. Finding 51.72 pts

    The quality gap widens in bad times.

    Over 1,294 months since 1919, Baa bonds yielded 1.72 points more than Aaa bonds in recession months, against 1.03 points in expansion months.[7]

  6. Finding 6+14%

    A guarantee can change behaviour fast.

    In January 2009, Hyundai launched an offer to take back financed or leased cars from buyers who lost their jobs. Its US sales that month rose 14% on a year earlier, while the whole US market fell 37%.[8]

Chapter 1 · 1 min

What a risk premium is

The idea has the same shape everywhere: whoever takes the risk sets the premium, and the premium grows with uncertainty. A bondholder charges a weaker company more interest. An investor wants a higher expected return from stocks than from government bonds. And a buyer wants a lower price, a smaller commitment or better terms from a supplier it has never used.

Terms used in this essay

TermMeaning
Risk premiumThe extra return demanded for bearing risk, over a safe alternative.
Equity risk premium (ERP)The extra return investors expect from stocks over a riskless investment.[1]
Implied equity risk premiumAn ERP backed out from current stock prices and expected cash flows, rather than from past returns.[4]
Default spreadThe extra interest a borrower pays over a riskless rate, set mainly by its credit rating.[2]
Country risk premiumThe extra equity premium Damodaran adds for investing in a given country, based on its sovereign rating.[3]
Credit spread (in our study)Moody's Baa corporate bond yield minus the 10-year Treasury yield.
Quality spread (in our study)Moody's Baa corporate bond yield minus the Aaa yield.
Buyer's risk premiumOur term for what a buyer charges an unproven supplier: discounts, delays, small first orders, slow payment.

How the premium shows up in finance and in buying

In financeIn B2B buying
Who takes the riskThe investor or lenderThe buyer, and the person inside who picked you
How it is chargedHigher yield or expected returnDiscount, longer decision, pilot order, longer payment terms, penalty clauses
What raises itWeak rating, unfamiliar country, a downturnNo track record, a new market, a downturn
What lowers itRating upgrade, collateral, guaranteesCheckable proof, guarantees, references, a visible track record
How visible it isQuoted dailyHidden inside each deal

Chapter 2 · 1 min

The evidence, part 1: what lenders charge for doubt

Damodaran publishes, free, a table linking a company's interest coverage to a synthetic credit rating and the default spread that goes with it, using data as of January 2026.[2] It is the plainest price list for doubt that we know of.

Figure 1. What a borrower pays over the riskless rate, by credit rating

Aaa/AAA
0.40%
Aa2/AA
0.55%
A1/A+
0.70%
A2/A
0.78%
A3/A-
0.89%
Baa2/BBB
1.11%
Ba1/BB+
1.38%
Ba2/BB
1.84%
B1/B+
2.75%
B2/B
3.21%
B3/B-
5.09%
Caa/CCC
8.85%
Ca2/CC
12.61%
C2/C
16.00%
D2/D
19.00%

Default spread (%), 0 to 20

Default spread over a riskless rate by synthetic rating, January 2026. Source: Damodaran, "Ratings, Interest Coverage Ratios and Default Spread," NYU Stern.

The curve is gentle at the top and steep at the bottom. Moving from AAA to BBB costs 0.71 points. Moving from BBB to B costs 2.10 points, and from B to CCC another 5.64. The less a lender can be sure of, the faster the price climbs.

Buyers work the same way, only without a published table. A supplier with a long, visible record sits near the top of the curve. A supplier the buyer has never heard of sits lower down, and pays for it in terms rather than interest.

Chapter 3 · 1 min

The evidence, part 2: what new markets cost

The same logic applies to geography. Damodaran estimates an equity risk premium for each country by adding a country risk premium, based on its sovereign bond rating, to a premium for a mature market.[3]

Figure 2. The extra return investors want, by market

Mature-market baseCountry risk premium
Germany: 4.23 + 0.00 = 4.23%
4.23
United States: 4.23 + 0.23 = 4.46%
4.23
United Kingdom: 4.23 + 0.78 = 5.01%
4.23
China: 4.23 + 0.91 = 5.14%
4.230.91
Poland: 4.23 + 1.10 = 5.33%
4.231.1
Mexico: 4.23 + 2.46 = 6.69%
4.232.46
Romania: 4.23 + 2.85 = 7.08%
4.232.85
India: 4.23 + 2.85 = 7.08%
4.232.85
Brazil: 4.23 + 3.24 = 7.47%
4.233.24
Vietnam: 4.23 + 3.90 = 8.13%
4.233.9
Nigeria: 4.23 + 8.41 = 12.64%
4.238.41

Equity risk premium (%), 0 to 13

Each label reads: mature-market base + country risk premium = total equity risk premium.

Country risk premium and total equity risk premium for selected markets. Source: Damodaran, "Country Default Spreads and Risk Premiums," NYU Stern, last updated 5 January 2026.

For a manufacturer planning to sell into a new country, this table is a reminder that capital already charges extra for less familiar markets. Buyers in those markets often face the reverse problem when they look at you. To a distributor in Mexico or Poland, a German or American supplier with no local references is the unfamiliar market. The premium they charge is the same idea pointed the other way.

Chapter 4 · 1 min

The evidence, part 3: when buyers cannot see quality

Why do buyers charge unknown suppliers at all? The economist George Akerlof gave the classic answer in 1970 in "The Market for 'Lemons'", published in the Quarterly Journal of Economics.[5] His example was used cars. Sellers know whether their car is good, buyers do not, so buyers price every car as if it might be bad, and owners of good cars stop selling.

The part of the paper most useful to a supplier is section IV, "Counteracting institutions." Guarantees come first. With guarantees, Akerlof wrote, "the risk is borne by the seller rather than by the buyer." Brand names come second: they "not only indicate quality but also give the consumer a means of retaliation if the quality does not meet expectations." His conclusion: "Informal unwritten guarantees are preconditions for trade and production."[5]

The investor Howard Marks adds the other half. In his 2014 memo "Risk Revisited," he warns that "if riskier investments could be counted on to produce higher returns, they wouldn't be riskier," and that when investors are risk-conscious, "they will demand generous risk premiums to compensate them for bearing risk."[6] Swap "investors" for "buyers" and you have the problem every new supplier faces. The Hive's entry on loss aversion explains why the fear of a bad choice weighs more than the hope of a good one.

If you are trying to win buyers in a market where nobody knows you yet, the Brave New audit looks at what a cautious prospect sees before deciding whether you are worth the risk.

Chapter 5 · 4 min

Original research: how the price of risk moves in a downturn

The tables above are snapshots. We wanted to know how much the price of risk moves over time, and when. So on 3 November 2026 we pulled every monthly reading of Moody's Baa and Aaa corporate bond yields since January 1919, the 10-year Treasury yield since April 1953, the official NBER recession dates, and the Federal Reserve's survey of bank lending standards since 1990, all from FRED.[7] We added Damodaran's implied equity risk premium for each year since 1961.[4] The full method is at the end of this essay.

Finding 1: the equity premium spikes in a crisis, then settles

Figure 3. The implied equity risk premium, selected years

Implied equity risk premiumAverage, 4.25%
246196119701979198819972006201520252008: 4.37% to 6.43%4.23%

Implied equity risk premium (%), end of year

Dashed gray line: 4.25% average.

End of yearImplied equity risk premium (%)
19653.32
19754.13
19796.45
19853.84
19953.29
19992.05
20054.08
20074.37
20086.43
20094.36
20156.12
20204.72
20254.23

Implied equity risk premium for the S&P 500 (free cash flow to equity model) at the end of each year. The full series for 1961 to 2025 is published with this essay. Source: Damodaran, "Historical Implied Equity Risk Premiums," NYU Stern; selection by Brave New Media.

Across 65 years, the implied premium averaged 4.25%. It was lowest at the end of 1999 (2.05%), at the height of the dot-com boom, and highest at the end of 1979 (6.45%). In 2008 it jumped from 4.37% to 6.43% in a single year, then fell back to 4.36% a year later.[4] Fear raised the price of risk fast, and calm brought it down again.

The equity premium does not track recessions neatly. Its average in years that included a recession month (4.46%, 14 years) was only slightly above the average in other years (4.19%). The bond market tells a clearer story.

Finding 2: credit spreads widened in every recession since 1953

Figure 4. What a solid company paid over the government, before and during each recession

Month beforePeak
0.93
1.23
1953
0.89
1.82
1957
0.92
1.32
1960
1.00
2.76
1970
1.69
3.31
1973
1.62
2.99
1980
1.89
3.82
1981
1.73
2.36
1990
2.95
3.34
2001
2.55
6.01
2008
2.11
3.47
2020

Spread (percentage points), by year the recession began

RecessionSpread the month before (points)Peak spread (points)Rise (%)
Aug 1953 to May 19540.931.2332
Sep 1957 to Apr 19580.891.82104
May 1960 to Feb 19610.921.3243
Jan 1970 to Nov 19701.002.76176
Dec 1973 to Mar 19751.693.3196
Feb 1980 to Jul 19801.622.9985
Aug 1981 to Nov 19821.893.82102
Aug 1990 to Mar 19911.732.3636
Apr 2001 to Nov 20012.953.3413
Jan 2008 to Jun 20092.556.01136
Mar 2020 to Apr 20202.113.4764

Moody's Baa corporate bond yield minus the 10-year Treasury yield, monthly. Peak is the highest month from the first recession month to six months after the last. Recession months are NBER-based. Source: Brave New Media analysis of FRED data (BAA, GS10, USREC), 3 November 2026.

Baa is the lowest investment-grade rating: solid companies, not distressed ones. Even so, in all 11 recessions since 1953 the extra yield they paid over the US government rose, by a median of 85%. The largest jump came in the 2008 crisis, when the spread more than doubled to 6.01 points in December 2008, the highest month in our data since 1953. The smallest came in 2001, when spreads were already high before the recession started.

Finding 3: the gap between strong and weak borrowers widens

Figure 5. The gap between strong and weaker borrowers, in good and bad times

ExpansionRecession
1.38
2.05
1919 to 1952 (285 / 123 months)
0.93
1.26
1953 to 1989 (367 / 77 months)
0.87
1.60
1990 to 2026 (406 / 36 months)
1.03
1.72
All years (1,058 / 236 months)

Baa minus Aaa (percentage points)

Months shown as expansion / recession.

Average monthly difference between Moody's Baa and Aaa corporate bond yields in NBER recession and expansion months. Source: Brave New Media analysis of FRED data (BAA, AAA, USREC), 3 November 2026.

This is the closest thing in public data to "what being less proven costs, and when". In every era, the extra yield a Baa borrower paid over an Aaa borrower was higher in recession months than in expansion months. Over the full 107 years, the gap averaged 1.72 points in recessions against 1.03 points in expansions. The record was 5.64 points in May 1932, in the depth of the Great Depression.

The lesson for a supplier: buyers' tolerance for an unproven name is not fixed. When their own business is under pressure, the distance between "the safe choice" and "the new choice" grows, just as it does between Aaa and Baa.

Finding 4: banks tighten fast, from a standing start

Figure 6. Banks tightening business lending, a year before and at the peak

A year beforePeak
24.6%
59.6%
2001
0%
83.6%
2007 to 2009
2.8%
71.2%
2020

Net share of banks tightening (%), 0 to 100

Net percentage of US banks tightening standards on commercial and industrial loans to large and middle-market firms, from the Federal Reserve's Senior Loan Officer Opinion Survey. "A year before" is four quarters before the first recession quarter. The survey series begins in 1990 Q2, so the 1990 to 1991 recession is not fully covered. Source: Brave New Media analysis of FRED data (DRTSCILM, USREC), 3 November 2026.

In early 2007, as many banks were easing standards as tightening them. By the fourth quarter of 2008, a net 83.6% were tightening. Around 2020 it went from 2.8% to 71.2% in six quarters. For a supplier, this matters twice. Your customers' credit gets harder to find, so they push for longer payment terms. And your own credit gets harder to find, at the same moment.

Limits of this study

Chapter 6 · 1 min

A framework: the Doubt Ledger

Every buyer keeps an informal ledger of doubt about a new supplier, and settles it through the terms of the deal. The framework below names each line, shows how the buyer charges for it, and gives the lever that shrinks it.

Source of doubtHow the buyer charges for itFinance parallelLever that shrinks itEvidence
"Can they deliver?"Discount, pilot orderLower credit rating, higher spreadCheckable proof: named results, references, published methodsSpreads rise steeply as ratings fall.[2]
"What if it goes wrong?"Penalty clauses, long decisionNo collateral or guaranteeA guarantee that moves risk to youGuarantees put risk on the seller (Akerlof); Hyundai's sales rose 14% in a falling market.[5][8]
"Have others trusted them?"Requests for many referencesShort credit historyVisible track record, case studies, brandBrand names give buyers "a means of retaliation".[5]
"Are they from here?"Slower approval in a new marketCountry risk premiumLocal proof, local language, local partnersEquity premiums range from 4.23% to 12.64% across markets.[3]
"Is now a bad time?"Longer payment terms, smaller ordersCredit spreads widen in recessionsFlexible first steps, staged paymentsBaa minus Treasury spreads rose in all 11 recessions since 1953.

Read the ledger from top to bottom when you prepare a proposal. Each row is a cost the buyer will charge you unless the proposal pays it down first.

Chapter 7 · 2 min

How to see, shrink and charge for the risk premium

Step 1: Measure the premium you are paying

Pull your last ten deals. Mark each one that came with a discount, a decision longer than your norm, a pilot or trial order, or payment slower than your terms. Then note whether the buyer knew you before. The difference between the two groups is the premium being unknown costs you.

Tip: Do the same split by market. If deals in a new country show more marks than deals at home, you are paying a local version of the country risk premium.

Step 2: Pay it down with proof the buyer can check

Claims cost nothing to make, so buyers ignore them. Proof that is hard to fake works better: named case studies, specific results, published methods, and references who will take a call. The Hive's entry on trust architecture breaks this into layers, and the Edelman Trust Barometer entry covers what trust surveys can and cannot tell you. You can see how we document results in the Brave New case studies.

Step 3: Move some of the risk onto yourself

A guarantee works because it moves risk from the buyer to the seller, exactly as Akerlof described.[5] It can be a free redo, a performance clause, or the right to cancel after a first phase.

Figure 7. A guarantee in a falling market

Hyundai
+14%
Entire US auto market
-37%

Change in US sales, Jan 2009 vs Jan 2008 (%)

Source: Knowledge at Wharton, "How Hyundai Sells More When Everyone Else Is Selling Less," 10 June 2009.

In January 2009, in the middle of the financial crisis, Hyundai offered to take back a financed or leased car if the buyer lost their job. That month its US sales rose 14% on a year earlier while the whole market fell 37%.[8] Other things were going on, so the 14% is not the guarantee's effect alone. But the offer removed the buyer's biggest fear at that moment, at a time when, as our data show, every price of risk in the economy was rising.

Step 4: Make the track record travel

Track record is the slowest lever and the strongest, but only if buyers can see it. Your brand is what you say about yourself; your reputation is what others believe based on what you have done, as the Hive's entry on brand versus reputation explains. Over time, reputation adds up to brand equity, the stored-up trust that lets you hold your price.

Keeping proof visible in every market you sell into is repetitive work, and it is where AI flows earn their place. One finished job can become a case note, a translated specification sheet and a proposal section for three markets, so the proof reaches more buyers without adding staff at the same rate.

Step 5: Price the risk you carry

You are a buyer of risk too. Every new customer might pay late, change the scope or disappear. Price that on purpose.

This case is made up, but it is typical. Say a small contract machine shop gets a $10,000 order from a new customer. From its own history, about one new customer in ten pays very late or not at all. That is about $1,000 of expected risk on the job. The shop can add about 10% to the price, ask for a 50% deposit, or split the order into batches paid before each ships. All three price the same risk. The deposit and the batches also protect cash.

Risk you carryPricing toolHow it works
Customer does not payDeposit or upfront paymentYou never carry the full amount
Scope keeps growingFixed scope with change ordersExtra work gets a new price
You cannot estimate the workTime and materials, or a fixed first phaseThe customer shares the estimate risk
Payment arrives lateShorter terms, staged invoicesYou are paid closer to when you work
A downturn arrivesReview terms when spreads widenYour terms move with the price of risk

Which to do first

ActionLedger lineEffortLikely impactWhen
Split your last ten deals by known vs unknown buyerAllLowHigh (shows the premium)This week
Add one checkable proof point to every proposalCan they deliver?LowHighThis month
Offer one guarantee you can afford to honourWhat if it goes wrong?MediumHighNext quarter
Translate and localise proof for each new marketAre they from here?MediumHigh in new marketsBefore entry
Set deposit and payment terms by customer riskYour own riskLowMediumNext contract

Chapter 8 · 1 min

What this means for leaders

For a manufacturer or supply-chain business, the buyer's risk premium is steepest exactly where growth lives: new customers, new channels and new countries. In each, you start lower on the buyer's ledger, and you pay for it in margin, time and cash.

Our data add a second point. The premium is not fixed. In every recession since 1953, the price lenders put on even solid companies rose, and the gap between strong and weaker names widened. Buyers behave the same way under pressure. A company planning expansion should expect the cost of being unknown to rise when the cycle turns, and should build its proof, guarantees and local presence while money is cheap and buyers are calm, not after.

The companies that do this well treat proof as an asset that compounds: every finished job, documented and translated, lowers the premium on the next one. If you want an AI flow that carries your proof to more buyers and new markets, start with the Brave New audit.

Chapter 9 · 1 min

Common mistakes

Chapter 10 · 1 min

Methodology

Data. From FRED, downloaded 3 November 2026: Moody's Seasoned Baa and Aaa Corporate Bond Yields (BAA, AAA, monthly from January 1919), the 10-year Treasury constant maturity yield (GS10, monthly from April 1953), the NBER-based recession indicator (USREC), and the net percentage of domestic banks tightening standards for commercial and industrial loans to large and middle-market firms (DRTSCILM, quarterly from 1990 Q2).[7] From Damodaran's site: historical implied equity risk premiums (histimpl.xls, 1961 to 2025), the ratings and default spread table (January 2026) and the country risk premium table (5 January 2026).[2][3][4] Downloads used a script that identified itself as "BraveNewResearch/1.0".

Measures. Credit spread is BAA minus GS10; quality spread is BAA minus AAA. A recession episode runs from the first to the last month flagged by USREC. For each episode we took the spread in the month before it began and the highest monthly spread from its first month to six months after its last. Recession and expansion averages use all months in each state. The 1918 to 1919 recession, already under way when the data begin, is excluded from Figure 4 and the episode list but its recession months count in Figure 5.

Denominators. 1,294 months (January 1919 to October 2026), of which 236 were recession months; 883 months with a Treasury comparison; 11 recessions since 1953; 65 years of implied equity risk premium; 147 quarters of the bank survey.

Limits. See "Limits of this study" above. Scripts and data are published with this essay.

Chapter 11 · 1 min

Frequently asked questions

What is a risk premium in simple terms?

A risk premium is the extra reward someone demands for taking a risk instead of a safe option. Investors want more from stocks than from government bonds, and lenders charge weaker borrowers more interest. In January 2026, a B-rated borrower paid a default spread of 3.21%, against 0.40% for a AAA borrower.

What is the equity risk premium today?

Damodaran's implied equity risk premium for the S&P 500 was 4.23% at the end of 2025. Over 1961 to 2025 it averaged 4.25%, with a low of 2.05% at the end of 1999 and a high of 6.45% at the end of 1979. It jumped from 4.37% to 6.43% during 2008.

Why do risk premiums rise in a recession?

Because fear rises and the chance of default rises with it. In our study, the gap between Baa corporate bond yields and 10-year Treasuries widened in all 11 US recessions since 1953, by a median of 85%. In 2008 it went from 2.55 to 6.01 percentage points.

How do buyers judge supplier risk?

Buyers usually cannot judge quality before they buy, so they look for proxies: track record, references, case studies, guarantees, and what others say. When those are weak, they charge a premium through discounts, long decisions, small trial orders and slow payment. Economist George Akerlof described this problem in 1970.

How do you price risk into a quote?

Estimate what could go wrong and how often, using your own past customers. Then cover it on purpose: a margin for uncertain work, a deposit against non-payment, change orders when scope grows, and shorter payment terms. A fixed first phase is a good way to price work you cannot yet estimate.

Your next step

This week, pull your last ten deals and mark every discount, long decision, trial order and late payment. Next to each, write the proof or guarantee that might have removed the buyer's doubt. Pick the one that appears most often and put it in your next proposal.

If you want to see how risky your offer looks from the buyer's side, in every market you sell to, start with the Brave New audit.

Brave New Media is a marketing agency that builds AI flows to help companies scale their ideas, grow their customer base and expand into new markets, with experience across supply chain and manufacturing.

See also

References

1. Damodaran, A. "Equity Risk Premiums: The 2012 Edition." 23 March 2012. aswathdamodaran.substack.com

2. Damodaran, A. "Ratings, Interest Coverage Ratios and Default Spread." Data as of January 2026. pages.stern.nyu.edu

3. Damodaran, A. "Country Default Spreads and Risk Premiums." Last updated 5 January 2026. pages.stern.nyu.edu

4. Damodaran, A. "Historical Implied Equity Risk Premiums" (histimpl.xls), 1960 to 2025. pages.stern.nyu.edu

5. Akerlof, G. A. "The Market for 'Lemons': Quality Uncertainty and the Market Mechanism." Quarterly Journal of Economics, 84(3), 1970, pp. 488 to 500. sfu.ca

6. Marks, H. "Risk Revisited." Memo to Oaktree clients, 3 September 2014. oaktreecapital.com

7. Federal Reserve Bank of St. Louis, FRED: series BAA, AAA, GS10, USREC and DRTSCILM. fred.stlouisfed.org

8. Knowledge at Wharton. "How Hyundai Sells More When Everyone Else Is Selling Less." 10 June 2009. knowledge.wharton.upenn.edu

9. Brave New Media. The price of risk across US recessions, 1919 to 2026, 3 November 2026. Data published with this essay.