Small Nasdaq stocks show the gap between a company lasting and shareholders profiting, while big-name slumps, Obamacare premiums and mortgage-bond buying test other assumptions.
Picked from 61 new articles across 16 sources · arrow keys for other days
Today in 30 seconds
30s left
One in six Nasdaq companies warns it may not last a year, including half of those under $50 million but only two of 1,589 above $2 billion. story 1 ↓
Of 145 Nasdaq companies warned in early 2025 over sub-$1 shares, 85 did reverse splits; the median survivor fell 47.3% while the S&P 500 rose 32.4%. story 2 ↓
Google, Intel and Nike sit deep below their highs, yet the S&P 500 is up 12% and two-thirds of its stocks have risen in 2026. story 3 ↓
Obamacare marketplace prices will rise 13% in 2027, with the cheapest high-deductible plan costing a 50-year-old about $700 a month; Tyler Cowen's mandate explanation faces timing questions. story 4 ↓
Fannie and Freddie added $11.7 billion in mortgage bonds in September, their biggest gain since March but still far from the promised $200 billion. story 5 ↓
1/5HumbleDollar· Daniel Oliveira· Nov 2
One in six Nasdaq companies doubts it can last a year
A count of US quarterly filings found 382 of 2,281 Nasdaq operating companies warning they may not survive 12 months, versus 3.4% on NYSE exchanges. The warnings cluster in tiny companies and rarely disappear.
Why it matters · Anyone buying small stocks can check for this disclosure in a minute, and it comes from management rather than outside critics; broad-index investors have far less exposure.
Know this first
Context from the editor, not the article · 27 sec
1 · The going-concern assessment
US accounting rules make management judge whether a company can meet its obligations for about a year after its financial statements are issued. If serious doubt remains after considering its plans, the company must say so.
2 · Why Nasdaq skews small
Nasdaq hosts many young, small companies alongside its giants, including lots of biotech firms. Its best-known names therefore give little sense of the typical listing, which can be tiny and dependent on outside funding.
3 · Staying alive by selling shares
A company burning cash can sell new stock to fund operations. This keeps it running, but each sale gives existing holders a smaller slice, and repeated sales at falling prices can steadily erode their investment.
2/5HumbleDollar· Daniel Oliveira· Nov 1
Most Nasdaq companies warned over sub-$1 shares survived; their shareholders still lost
Daniel Oliveira tracked 145 Nasdaq companies warned for trading under $1. About three-quarters stayed listed through October 2026, yet their median share price fell by nearly half while the broader market climbed.
Why it matters · If you own a cheap stock, headlines about a company "regaining compliance" can sound like recovery; this data separates the listing's survival from your investment's.
Know this first
Context from the editor, not the article · 31 sec
1 · Why exchanges set minimum prices
Stock exchanges set listing standards, including a minimum share price, to keep very weak companies off their main markets. Falling short does not end trading immediately. The company is warned first and given time to fix the problem.
2 · What a reverse split does
A reverse split swaps many old shares for fewer new ones at a proportionally higher price. The company's total market value and each holder's stake stay the same on the day, so only the quoted price changes.
3 · What delisting means
A delisted stock leaves its exchange. It may still trade over the counter, a less regulated dealer market, but typically with less liquidity and less scrutiny. Deregistration goes further, ending a company's regular public reporting to the US Securities and Exchange Commission.
3/5A Wealth of Common Sense· Ben Carlson· Nov 2
Big-Name Stocks Are Tumbling, but the S&P 500 Is Still Up 12%
Many famous stocks have fallen 30% to 79% from their highs, Ben Carlson shows, yet the S&P 500 is up almost 12% this year. Stock-pickers feel the pain; index fund owners barely notice.
Why it matters · Headlines about plunging famous stocks can push investors to react; these numbers show how little single-stock crashes have so far meant for a diversified portfolio.
Know this first
Context from the editor, not the article · 18 sec
1 · How the S&P 500 works
The S&P 500 tracks about 500 large US companies, weighted by market value, so bigger companies move it more. An index fund simply buys those stocks in roughly those proportions, rather than choosing winners.
2 · Two ways to measure returns
A drawdown is the fall from a stock's highest price. A year-to-date return is the change since January 1. The same stock can look terrible on the first measure and fine on the second.
S&P 500 returns for 2026 compared with the average stock (almost 15%) and median stock (around 9%).1 / 2
Google's stock fell 12% the day after it reported earnings last week, one of the worst days in the company's history. It now sits more than 30% below its high. The article doesn't say what in the report troubled investors.
4/5Marginal Revolution· Tyler Cowen· Nov 1
Obamacare marketplace prices will rise 13% in 2027, renewing doubts about stability
Average Obamacare marketplace premiums will rise 13% next year, KFF, a health research group, calculates. Economist Tyler Cowen argues weak mandate enforcement drives healthy buyers away, making the markets unstable. Subsidies and timing complicate his case.
Why it matters · Anyone who is self-employed, retiring early or between jobs may rely on these markets, and whether they erode or collapse shapes both their options and federal spending.
Know this first
Context from the editor, not the article · 28 sec
1 · What the marketplaces are
The Affordable Care Act, or Obamacare, created regulated markets where people without job-based coverage buy insurance. Insurers cannot reject or charge more for pre-existing conditions, and income-based tax credits lower premiums for many buyers.
2 · The individual mandate
Because insurers must accept sick people, the law also penalized most people who stayed uninsured, pushing healthy people into the pool. A 2017 tax law, signed by President Trump, cut the federal penalty to $0 starting in 2019.
3 · The enhanced subsidies
5/5Calculated Risk· CalculatedRisk by Bill McBride· Nov 1
Fannie and Freddie Buy More Mortgage Bonds, Still Far From $200 Billion
Fannie Mae and Freddie Mac increased mortgage-bond holdings in September, but their $50.7 billion gain this year is about a quarter of the $200 billion President Trump implied. Markets have barely reacted, economist Tom Lawler reports.
Why it matters · Government-controlled buying of mortgage bonds is one lever on mortgage rates, so tracking actual purchases against political promises shows how much support the housing market is really getting.
Know this first
Context from the editor, not the article · 31 sec
1 · Fannie, Freddie and mortgage bonds
Fannie Mae and Freddie Mac buy home loans from lenders, bundle them into mortgage-backed securities and guarantee them against borrower default. They can also keep those bonds as investments. The government has controlled both companies through conservatorship since 2008.
2 · Bond demand and mortgage rates
Lenders price new mortgages largely off what investors pay for mortgage bonds. More demand lifts bond prices and lowers yields, which can pull mortgage rates down. The extra yield over Treasuries, called the spread, shows what investors demand for the added risk.
A count of US corporate filings finds that 382 of the 2,281 operating companies listed on Nasdaq, about one in six, warn in their latest quarterly report that they may not survive the next year. On the New York Stock Exchange and 1, its smaller-company sibling exchange, the rate is 3.4%.
Daniel Oliveira reported the tally on the HumbleDollar forum. It covers every US-listed company that filed a quarterly report between July and late October, using Securities and Exchange Commission (SEC) filings processed by Equibles, a data provider.
What the warning means
The key phrase is that conditions "raise substantial doubt" about a company's ability to continue as a 2, the accounting term for a business expected to keep operating and paying its bills. The doubt comes from management itself, not from a 3, an investor who profits if the price falls.
Under US accounting rules, a rescue plan removes the doubt only if it will probably happen and work. Shares a company merely hopes to sell generally don't clear that bar, so even a serial fundraiser can keep printing the warning.
Mostly a small-company problem
Oliveira attributes most of the gap between exchanges to size. Half of listed companies worth under $50 million carry the warning. Of the 1,589 worth more than $2 billion, only two do. Drug developers are the largest group, at a third of the total, which fits a business that often has no product revenue and lives on funding rounds.
Since the count is of companies rather than dollars, one in six overstates what a broad, 4 index fund holds, though the post gives no value-weighted figure.
It rarely goes away
Oliveira was most surprised by the persistence. Of companies that warned in a report filed between January and May, 94% warned again in their next one. And 36% of the Nasdaq companies with the warning traded below $1, the minimum bid Nasdaq generally requires for continued listing.
Not a bankruptcy forecast
Many of these companies raise money and carry on, Oliveira notes, usually by selling new shares to existing investors at lower prices. That can still hurt shareholders. An illustrative example, not from the post: a company with 10 million shares at $2 sells 10 million more at $1. A 1% holder now owns 0.5%, and the price would settle near $1.50 if the market values the company at its old worth plus the new cash.
The one-minute check
His advice for small-company buyers is to search the latest 5, the quarterly report filed with the SEC, for "going concern". He argues it says more than most press releases.
The post gives no bankruptcy rates, shareholder returns or historical trend, so it shows how common the warning is, not how well it predicts failure.
A count of US corporate filings finds that 382 of the 2,281 operating companies listed on Nasdaq, about one in six, warn in their latest quarterly report that they may not survive the next year. On the New York Stock Exchange and 1, its smaller-company sibling exchange, the rate is 3.4%.
Daniel Oliveira reported the numbers in a post on the HumbleDollar forum. The count covers the latest quarterly report of every US-listed company that filed one between July and late October, drawn from Securities and Exchange Commission (SEC) filings and processed by Equibles, a data provider. Oliveira says he expected the warning to be rare.
What the sentence means
The phrase to look for is that conditions "raise substantial doubt" about a company's ability to continue as a 2. "Going concern" is the accounting term for a business expected to keep operating and meeting its obligations.
The source of the warning gives it its weight. As Oliveira puts it, this is management itself, not a 3 (an investor who profits if the price falls) or a critic, looking at the cash and the debts and being unable to say the bills will be paid for the next 12 months.
The accounting test helps explain how the sentence behaves. Under US rules, management first asks whether conditions raise substantial doubt, then whether its plans probably fix the problem. A plan counts only if it is probable that it will be carried out and will work. A company hoping to sell shares it has not yet sold, with no committed buyers, generally struggles to clear that bar. So a business can have a long record of raising money and still have to print the warning.
Where the warnings sit
Oliveira attributes most of the gap between the exchanges to size, and the extremes are stark:
Under $50 million in market value: half of listed companies carry the warning.
Over $2 billion: just two of 1,589 companies do.
By industry: drug developers are the largest group, at a third of the total.
The drug-developer figure fits how that business generally works. Many have no product to sell, spend heavily on clinical trials, and depend on repeated funding rounds until a drug succeeds or fails. For them, a year of uncertain cash can be a normal state rather than a sudden crisis.
The size split also shows what the headline does and does not mean. One in six counts companies, not dollars. Since the warning is almost absent among large companies, it probably covers a very small share of Nasdaq's total market value, so an investor in a 4 index fund has far less exposure than "one in six" suggests. That is an inference from the size data; the post gives no value-weighted figure.
The warning tends to stay
Oliveira says the persistence surprised him most. Of the companies that warned in a report filed between January and May, 94% warned again in their next report. Among Nasdaq companies with the warning, 36% traded below $1.
That price matters on its own. Nasdaq's listing rules generally require a minimum bid of $1, so a stock stuck below it can face delisting unless the price recovers, often through a 5, which merges shares to lift the per-share price without adding any value. The post does not discuss this, but it means many of these companies face a listing problem on top of a cash problem.
Survival isn't the same as a good outcome
Oliveira is explicit that the warning is not a bankruptcy forecast. Many of these companies raise money and carry on, usually by selling new shares to existing investors at lower prices.
That route can keep the company alive while still costing shareholders, through 6. An illustrative example, not from the post: a company has 10 million shares at $2, worth $20 million, and you own 100,000 shares, or 1%. It sells 10 million new shares at $1, raising $10 million. Your stake falls to 0.5%. If the market values the company at its old $20 million plus the new cash, the share price settles near $1.50, a 25% loss in a transaction that "saved" the business. Repeat that a few times and a company can survive for years while its early shareholders lose most of their money.
The practical check
Oliveira's suggestion is simple: before buying a small company, search its latest 7, the quarterly report filed with the SEC, for "going concern". He argues that this minute of work says more than most press releases. SEC filings are free to read through the agency's EDGAR database.
What the numbers don't show
The post is a short summary, and Oliveira points to a fuller write-up on Equibles' site for the method and breakdowns. Several questions stay open in the post itself:
Outcomes: it gives no rate of bankruptcy, delisting or shareholder returns for companies that warn, so it does not show how much the warning predicts.
Trend: it is a single snapshot, so it cannot say whether one in six is high or low by historical standards.
Persistence: the 94% describes companies' next reports. The post does not say how it treats companies that stopped filing, perhaps because they failed, or what happened to the 6% that dropped the warning.
Causation: smaller companies warn far more often, so the warning may partly be a marker of size and early-stage funding rather than a separate signal.
A search for the phrase is therefore a screen, not a verdict. Its clearest use is to tell you that a company's survival depends on money it does not yet have.
Remember
Treat the warning as a screen for dependence on future funding, not as a prediction of bankruptcy.
A company can survive repeated share sales while its existing shareholders still lose much of their investment.
Counts of companies can mislead: warnings cluster among tiny firms that carry little weight in value-weighted indexes.
Daniel Oliveira, posting on the personal-finance site HumbleDollar, followed 145 companies after 1 warned them that their shares had become too cheap. The warning rarely ended a listing, but shareholders in the companies that stayed listed still typically lost close to half their money.
The rule
Nasdaq, a major US stock exchange, sends a warning letter when a stock closes under $1 for 30 trading days in a row. The company then gets 180 days to get back above $1. Oliveira tracked companies that disclosed a first letter in the first half of 2025, checking on them on October 26, 2026.
Who survived
108 (74.5%) still traded on a US exchange.
19 were 2 (removed from the exchange), suspended or 3 (no longer filing regular public reports).
7 went bankrupt.
11 were acquired.
Why survival wasn't a win
Of the 145, 85 used a 4, typically 1-for-20, which swaps many old shares for fewer new ones. As an illustration, 100 shares at $0.40 become 5 shares at $8. Both holdings are worth $40. The price clears the exchange's bar, but the business hasn't changed.
Returns, measured from the day each letter was disclosed, show the gap:
The 5 stock still listed, meaning the one in the middle of the group, fell 47.3%. The 6, an index of large US companies, gained 32.4% over the same windows.
Companies that were delisted or went bankrupt lost a median 97.7%.
In illustrative terms, $1,000 became about $527 in the median survivor, about $1,324 in the index and about $23 in the median failure.
Oliveira's conclusion is that "regaining compliance" is a statement about the listing, not about your money.
The sharper warning sign
Some companies had a second listing problem on top of the low price, such as too little 7 (assets minus liabilities) or a late financial filing. They were more than twice as likely to end up delisted or bankrupt: 27.6% against 11.5%.
What the numbers don't show
This doesn't prove the letters caused the losses. A stock usually falls below $1 because the business is already struggling, so the letter may just mark a decline under way.
The S&P 500 is also an imperfect yardstick for small, distressed companies. And this is a single group followed for less than two years. The post doesn't test whether selling on receipt of the letter beat holding, and it leaves the method to a longer study.
The practical reading is narrower than "sell on the letter": don't treat a reverse split or restored compliance as evidence that the company has recovered. Check its profits, its cash and whether a second listing problem has appeared.
Daniel Oliveira, posting on the personal-finance site HumbleDollar, followed 145 companies after 1 warned them that their shares had become too cheap. He found that the warning rarely ended a company's listing, but shareholders in the companies that stayed listed still typically lost close to half their money. His conclusion is that "regaining compliance" describes the listing, not your investment.
The rule and the sample
Nasdaq, one of the two main US stock exchanges, requires listed shares to hold a minimum price of $1. When a stock closes under $1 for 30 trading days in a row, the exchange sends a warning letter and the company gets 180 days to get back above the line.
Oliveira took every company he identified as disclosing a first letter in the first half of 2025, then checked where each stood on October 26, 2026. That gives each company roughly 16 to 22 months of follow-up.
Who survived
Outcome by October 26, 2026
Companies
Share of 145
Still trading on a US exchange
108
74.5%
Delisted, suspended or 2
19
13.1%
Bankrupt
7
4.8%
Acquired
11
7.6%
Delisted means removed from the exchange. Deregistered means the company has also ended its regular public reporting. Taken alone, the table looks reassuring: only about 18% of the group, 26 companies, ended in delisting, suspension, deregistration or bankruptcy.
How companies stayed listed
The main tool was the 3, which swaps many old shares for fewer new ones. Of the 145 companies, 85 did one, typically at a ratio of 1-for-20.
The arithmetic shows why this solves the exchange's problem and not the investor's. As an illustration, suppose you own 100 shares at $0.40, worth $40. After a 1-for-20 split you own 5 shares at $8, still worth $40. The company is now comfortably above $1, while its revenue, debts and prospects are exactly what they were the day before.
Oliveira puts it bluntly: a reverse split fixes the share price on paper and nothing else.
What shareholders got
The returns run from the day each letter was disclosed:
Still listed: the 4 stock, meaning the one in the middle of the group, fell 47.3%. The 5, an index of large US companies, gained 32.4% over the same windows.
Delisted or bankrupt: the median loss was 97.7%.
In illustrative dollar terms, $1,000 would have become about $527 in the median surviving stock, about $1,324 in the index, and about $23 in the median failed company. The post does not report returns for the 11 acquired companies.
The 47.3% figure also comes with a built-in flattering effect. It covers only companies that stayed listed, so it leaves out the wipeouts.
Two problems are worse than one
Some companies had the low price plus a second listing issue, such as too little 6 (assets minus liabilities) or a late financial filing. These were more than twice as likely to end up 7 or bankrupt: 27.6% against 11.5%.
The post does not give the group sizes. If those rates cover all 145 companies, they are consistent with about 16 failures among 58 companies with two problems and 10 among 87 with one. That is my arithmetic rather than the author's, but it shows the comparison rests on small counts. It is a useful warning sign, not a precise probability.
What to be careful about
Oliveira's numbers describe what followed the letters. They do not show that the letters caused the losses. A stock under $1 has usually already fallen because the business is in trouble, so the letter may simply mark a decline already under way.
Several other limits apply:
The benchmark is an imperfect match. The S&P 500 holds large, established companies, while these are distressed small ones. A small-company index would be a fairer test, although the gap is wide enough that the broad direction seems unlikely to depend on it.
Medians hide the spread. The middle stock fell by about half, but the post doesn't report how many stocks rose or whether a few large winners changed the group's average return.
The time period is short and specific. This is one cohort over less than two years, during a strong market for the S&P 500.
The method is elsewhere. The post points to a longer study with the method and source filings. It doesn't explain here how the returns were calculated.
The rules may matter. Nasdaq can give some companies a second 180-day period, and in January 2025 it tightened limits on long non-compliance and repeated reverse splits. The post doesn't discuss whether this shaped the outcomes.
The post also leaves the most practical question open: whether selling at the letter beat holding. The median survivor's decline points that way, but Oliveira doesn't test it as a strategy, and he ends by asking readers whether they sold or waited.
For an individual investor, the usable reading is narrower than "sell on the letter." Treat a compliance notice or a reverse split as neutral at best about the business. Then look at what actually determines value: whether the company is making money, how much cash it has left, and whether a second listing problem has appeared.
Remember
A reverse split changes the quoted share price, not the company's value or the health of its business.
A second listing problem alongside a sub-$1 price was the clearer danger signal: 27.6% versus 11.5% ending delisted or bankrupt.
Survivor figures flatter results: the 47.3% median decline excludes the delisted and bankrupt companies, whose median loss was 97.7%.
Google has plenty of company. Investment writer Ben Carlson lists the damage as 1, the percentage fall from a stock's peak:
Chips and AI: Intel lost almost 30% on Friday alone and is 54% off its high. 2, which rents out computing power for AI work, is down 60% since May. Sandisk and Western Digital are each down 51%.
Other tech: Oracle is down 60%, Coinbase more than 60%, and Netflix more than 50%.
Consumer names: Nike is down 79%, Chipotle 55%, Pepsi 32%, and Home Depot 25%.
The index tells a different story
Yet the 3, an index of about 500 large US companies, is up almost 12% in 2026. Through Friday's close, two-thirds of its stocks were up, and nearly half had gained 10% or more. More were up 20% or 30% than down by those amounts.
The average stock has gained almost 15% and the 4, the one in the middle of the ranking, around 9%. An average above the median usually means a smaller group of big winners is pulling it up. So the typical company has had a decent year rather than a spectacular one, but hardly a bloodbath.
How both can be true
First, the two sets of numbers start from different places. Drawdowns count from a stock's peak, while 2026 returns count from January 1. Many of these stocks rose sharply before falling. As an illustration, a stock that goes from $100 to $200 and then drops 40% lands at $120: a painful slide, yet still up 20% on the year.
Second, an index dilutes individual disasters. A stock making up 5% of an index that falls 12% (illustrative numbers) takes only 0.6 percentage points off the index. Someone who holds that stock as a big share of their portfolio takes most of the blow.
What Carlson concludes
Carlson argues this is the case for 5. Stock-pickers in the wrong corners of the market may feel sick right now, while someone who owns the whole market "would never know" about the turmoil.
He adds a caveat: the sell-off could be a precursor, with the wider market following these stocks down. An index fund owns the losers too. It has been cushioned so far because other stocks have kept rising, and that protection would shrink if most stocks fell together.
Google's stock fell 12% the day after it reported earnings last week, one of the worst days in the company's history. It now trades more than 30% below its high. The article doesn't say what in the report troubled investors, only that the numbers were big and the market didn't like what it heard.
Investment writer Ben Carlson uses that drop as the starting point for a tally of the damage across the market, and then sets it against the numbers for the market as a whole. The two pictures look as if they describe different markets.
The damage
Carlson measures the losses as 1, the percentage fall from a stock's peak price:
AI-linked giants: Meta is 20% off its high, after a 40% run-up in a month. Amazon is down more than 15% and Nvidia nearly 15%.
Single-day collapses: Intel dropped almost 30% on Friday alone and is now 54% below its all-time high. 2, which rents out computing power for AI work, fell 10% the same day and is down 60% since May.
Chips and storage: Sandisk and Western Digital are each down 51%, Seagate Technology 41%, and Micron 27%.
Other tech: Applied Materials is down 46%, Broadcom 32%, and Dell 28%.
Other well-known technology-driven companies: Netflix is down more than 50%, Coinbase more than 60%, Oracle 60%, and Uber 35%.
Consumer names: from their all-time highs, Nike is down 79%, Chipotle 55%, Pepsi 32%, and Home Depot 25%.
The market underneath
The 3, an index of about 500 large US companies, is up almost 12% in 2026. Carlson's breakdown, through Friday's close, shows the gains are broad:
Two-thirds of the stocks in the index are up this year.
Nearly half are up 10% or more.
More stocks are up 20% or 30% than are down by those amounts.
The average stock is up almost 15%, and the 4 stock around 9%.
The median is the stock in the middle of the ranking. An average above the median usually means a smaller group of big winners is pulling the average up. So the typical company has had a decent year, while a minority has had an excellent one.
One further inference, ours rather than Carlson's: the S&P 500 weights companies by market value. If the "average stock" figure is a simple average, then the index's 12% falling short of that 15% suggests the largest companies, as a group, have lagged the average stock this year.
How both pictures can be true
The first reason is that the two sets of numbers use different starting points. A drawdown starts at the peak, while a 2026 return starts on January 1. As Carlson notes, many of these stocks had huge gains before they fell. Consider an illustrative stock that starts the year at $100, runs to $200, and falls 40% to $120. Its owner has watched a painful decline, yet the stock is up 20% for the year. Some of the all-time highs may also be old, and the article gives no dates for them, so these losses didn't necessarily all happen in 2026.
The second reason is diversification. An index spreads its money so widely that one company's collapse is diluted by hundreds of others. As an illustration, a stock making up 5% of an index that drops 12% subtracts only 0.6 percentage points from the index that day. Meanwhile, someone holding that stock as a large share of a small portfolio takes most of the hit.
The arithmetic also explains why these losses are so hard for concentrated investors. A stock that falls 50% has to double to get back to even. A 79% fall, like Nike's, requires a gain of nearly 380%.
Carlson's argument
Carlson argues this is what makes the market "equal parts exciting and excruciating." Pick the right stocks and it's wonderful; pick the wrong ones and you can be "in a world of pain even when the overall market is doing just fine." He presents that as the appeal of 5: if you own the whole market, you would never know about the turmoil beneath the surface.
Research outside the article supports the underlying idea. Finance professor Hendrik Bessembinder found that, from 1926 to 2016, the best-performing 4% of US-listed stocks accounted for all of the stock market's net wealth creation above 6. Owning everything guarantees you hold those rare winners. Picking a few stocks makes it easy to miss them.
What this doesn't show
This is one snapshot from one strong year, and Carlson says so himself: the turmoil may be "a precursor of things to come," with the wider market following the fallen stocks down. He makes no prediction either way.
"You would never know" also deserves a qualifier. Index funds own the losers too. Their shareholders are insulated only because other stocks have been rising enough to offset the falls. And since the S&P 500 is weighted by size, heavy simultaneous declines in its largest members would hit it much harder than an equal number of drops among smaller companies. Diversification has muted this round of losses, but it wouldn't prevent a loss if most stocks fell together.
Remember
Check the starting point of any loss figure: from-the-peak and year-to-date numbers can tell opposite stories about one stock.
Recovery math is brutal: a 50% loss needs a 100% gain, and a 79% loss needs nearly 380%.
Diversification dilutes single-stock disasters, but it can't prevent losses if most of the market falls together.
From 2021 through 2025, temporary federal credits made coverage much cheaper and extended help to higher earners. They lapsed at the end of 2025, sharply raising what many buyers pay, though Cowen's post does not discuss them.
Image: Marginal Revolution
Average premiums in the Obamacare marketplaces will rise 13 percent in 2027, according to 1, a health research group. It combined new prices the Trump administration published for the 30 states with federally run markets with figures from the other states.
The New York Times report on the analysis gives a concrete price: the cheapest 2, a high-deductible option, will average about $700 a month for a single 50-year-old. That is $8,400 a year before the plan pays for most care.
Cowen's argument
Economist Tyler Cowen, writing on Marginal Revolution, says the markets are proving unstable, as he expected. Higher prices make people who expect to stay healthy even less likely to sign up, and weak mandate enforcement gives them little reason to buy.
The worry is 3: when buying is optional, those who expect big medical bills are the most likely to buy. An illustrative example: 80 healthy people cost $2,000 a year each and 20 sick people cost $20,000 each, so the average is $5,600. If 20 healthy people leave, the average for those remaining jumps to $6,500, a 16% rise. That increase can push out more healthy buyers. Taken to the extreme, this is a "4."
Cowen also makes a political argument. When the 5, the requirement to hold coverage, carried real penalties, the law was very unpopular and drew protests. He says its costs fell on an identifiable minority, those who bought only because they had to, while its benefits were spread across a scattered majority. By relaxing enforcement, he argues, Trump saved Obamacare politically for a while but created this longer-term problem. He adds that Democrats rarely boast about the law.
What complicates the story
The post is a short opinion piece and doesn't test this claim. Three facts outside it matter:
The timing is awkward. The federal penalty fell to $0 in 2019, yet enrollment later reached a record of about 24 million in 2025. Generous subsidies, not a mandate, drew people in.
Subsidies blunt a spiral.6 generally cap what eligible buyers pay as a share of income and grow when premiums rise. Those buyers are partly shielded, while people earning too much for credits face the full price.
Not all of the 13% is selection. Rising medical costs also lift premiums, and the excerpt gives no breakdown of the causes.
A fairer reading may be that the markets depend heavily on federal money. When subsidies shrink, they lose healthy buyers. Whether that becomes slow erosion or Cowen's instability will show up in 2027 enrollment and in the health of those who stay.
Average premiums in the Obamacare marketplaces will rise 13 percent in 2027, according to an analysis by 1, a health research group. The Trump administration published 2027 prices for the 30 states whose markets the federal government manages. KFF combined that data with state figures to calculate the national increase.
The New York Times report on the analysis gives one price that shows the scale. The cheapest 2, a high-deductible option, will average about $700 a month for a single 50-year-old, a figure attributed to a Ms. Cox in the excerpt. That is $8,400 a year. Bronze plans are designed to cover only about 60% of an average person's medical costs, so a buyer can pay that premium and still face large bills before coverage does much.
For someone with no subsidy, the burden is easy to see. At an illustrative income of $65,000, $8,400 is about 13% of gross pay, before any deductible.
Cowen's argument
Tyler Cowen, the economist who wrote the post on Marginal Revolution, says the exchanges "seem not to be stable," as he predicted. Rising prices, he writes, make people who expect to be healthy even less likely to sign up, because enforcement of the 3, the requirement to hold coverage, has become quite weak.
The economics here are standard. The law requires insurers to accept everyone at prices that don't depend on health. That is good for sick people, but it creates 4: when buying is optional, the people most likely to buy are those who expect to use the coverage.
An illustrative example shows how this feeds on itself. Suppose a pool has 100 people. Eighty are healthy and cost $2,000 a year each, and 20 are sick and cost $20,000 each. Total costs are $560,000, so the break-even premium is $5,600. Now 20 healthy people decide that isn't worth it. The remaining 80 people cost $520,000, and the premium must rise to $6,500, about 16% more. No one's health has changed, but the higher price gives more healthy people a reason to leave. If that loop runs to its end, it is called a 5: premiums climb until the market holds mostly high-cost buyers or insurers withdraw.
The political trap
Cowen's more distinctive point is political. He argues that when the mandate was tightly enforced with real penalties, the ACA was very unpopular and led to significant protests. In his account, the law's benefits went to a broad, scattered majority, while its costs landed on a well-identified minority: those who signed up only because the mandate made them.
So he sees a trade-off. Trump, in his first term, relaxed mandate enforcement, which by Cowen's reckoning politically saved Obamacare for a while but created the longer-term problem now appearing. Strictly, Congress set the federal penalty to $0 from 2019 through the 2017 tax law, which Trump signed. A few states have kept mandates of their own.
Cowen closes with an observation, not evidence: he rarely hears Democrats boasting about Obamacare, and says they tend to talk as if little or nothing has been done for the uninsured.
Where the argument needs more evidence
The post runs to a few short paragraphs and offers no enrollment or risk-pool data. Its headline claim is a judgment, and three pieces of background bear on it.
The timing doesn't fit neatly. The penalty disappeared in 2019. In 2017, the 6, Congress's nonpartisan scorekeeper, projected that repeal would raise premiums by about 10% in most years and increase the uninsured substantially. Yet marketplace enrollment went on to reach a record of about 24 million in 2025. A weak mandate alone did not unravel these markets. Enrollment rose after temporary enhanced subsidies, in place from 2021 through 2025, made coverage far cheaper. They lapsed at the end of 2025. Cowen doesn't mention them, but they may explain the current strain better than a mandate that vanished years ago.
Subsidies work against a classic spiral.7 are pegged to the price of a benchmark 8 and limit what eligible buyers pay for it as a share of income. When premiums rise, the credit generally rises too, and the federal government absorbs much of the increase. Those buyers don't feel the full price signal, so they have less reason to leave. People with incomes above the cutoff, now a bit over $60,000 for a single person, pay the full price. They are the group most exposed to the loop described above, and the likely result is uneven erosion, not a market-wide collapse. That is my inference from how the credits work, not a finding in the post.
A 13% rise isn't a measure of selection. Premiums also rise when hospital prices, drug spending and use of care go up, and those pressures affect job-based insurance too. The excerpt doesn't say how much of the increase comes from a sicker pool. Nor does a sticker-price average show what households actually pay after credits.
What would settle it
The useful evidence will arrive in 2027: how many people enroll and pay their premiums, whether younger and healthier buyers leave disproportionately, whether insurers exit, and what share of next year's rate increases insurers attribute to a worsening risk pool. If those measures deteriorate together, Cowen's claim gets much stronger. If enrollment shrinks but stabilizes among subsidized buyers, the better description may be a market that survives on federal money and shrinks when that money is cut back.
The post doesn't give the prior-year price of the bronze plan, state-by-state variation or KFF's method beyond the sources it combined. Those details would need to come from KFF's analysis or the Times report.
Remember
Unsubsidized buyers carry the sharpest risk: subsidies shield eligible households from much of a sticker-price increase.
Enrollment hit a record years after the federal penalty ended, so a weak mandate alone can't explain today's strain.
Watch who enrolls in 2027, not just average premiums; the mix of healthy and sick buyers tests stability.
State and local governments issue municipal bonds, whose interest is generally exempt from federal income tax. That tax break usually lets them pay lower yields than Treasuries, so investors compare the two markets using a ratio of their yields.
Summary statistics from the Fannie Mae and Freddie Mac monthly volume reports discussed by Lawler.1 / 3
Fannie Mae and Freddie Mac, the government-controlled mortgage companies known as 1, added to their mortgage-bond holdings in September. Housing economist Tom Lawler's review of their monthly reports shows the buying remains far below what the White House signaled in January.
Holdings of 2, mortgage bonds guaranteed by a government-linked agency, rose $9.923 billion at Fannie and $1.785 billion at Freddie. The combined $11.708 billion was the largest monthly gain since March.
Short of the $200 billion
Combined holdings were still little changed from April, and they are up only $50.663 billion this year. That is about a quarter of the $200 billion implied by President Trump's January 8 post. The idea behind such buying is that extra demand lowers mortgage-bond yields, which can pull mortgage rates down.
After a report on falling holdings, Bill Pulte, head of the companies' regulator, the 3, posted on September 18 that they were starting to buy "large quantities." Lawler notes that purchases committed that late would almost certainly settle in October or later, so they wouldn't appear in September's data. Later reports will show whether that buying happened.
Meanwhile, both companies' interest-rate risk stayed historically elevated, as measured by the estimated loss in market value if rates rose 50 4, or half a percentage point.
Yields fell, spreads didn't
The yield on 5, a benchmark for bonds backed by new mortgages, fell 34 basis points in October. That reversed exactly half of September's 68-point jump. Oddly, the drop gained speed just after somewhat hotter-than-expected inflation reports.
The summer's strong link between interest rates and oil prices also vanished, replaced by a link between rates and tech stocks. Several analysts read that as investors moving from tech into bonds.
The spread between CCMBS and Treasury yields still rose modestly. Lawler partly blames higher 6, the size of future rate swings priced into options. Bigger swings make mortgage bonds riskier because homeowners refinance when rates fall and stay put when rates rise, both of which hurt bondholders. So far, news of the extra purchases has had little market impact.
Munis bounce, partly
Municipal bonds had what Lawler calls a horrific September. Top-rated yields rose much more than Treasury yields, driven by a huge spike in selling by individuals. Some of that was 7, selling at a loss to offset capital gains elsewhere.
The 8, a muni's yield divided by a comparable Treasury's, shot up, signaling cheaper munis. In October's second half, muni yields fell further than Treasury yields. Ratios eased, but stayed above August's level and well above those from earlier in the year.
Fannie Mae and Freddie Mac, the government-controlled mortgage companies known as 1, added to their mortgage-bond holdings in September. Their purchases are still well short of what the White House signaled in January, according to housing economist Tom Lawler's review of their latest 2, the reports each company publishes on its portfolio and business activity.
Fannie's holdings of 3, meaning mortgage bonds guaranteed by a government-linked agency, rose $9.923 billion in the month. Freddie's rose $1.785 billion. The combined $11.708 billion gain was the largest since March.
The gap between the promise and the purchases
The longer view is less impressive. Combined holdings at the end of September were little changed from April, because they fell during the months in between. Since the end of 2025, Fannie is up $38.203 billion and Freddie $12.46 billion, for a total of $50.663 billion. That is about a quarter of the $200 billion in purchases implied by President Trump's January 8 post.
The logic of such buying is straightforward: extra demand for mortgage bonds should raise their prices and lower their yields, which can feed through to lower mortgage rates. The drop in holdings over recent months surprised many market analysts, Lawler writes.
After a story reported the April-to-August decline, Bill Pulte, director of the 4, the companies' regulator, posted on September 18 that "(w)e (the GSEs) are beginning to buy even more, large quantities (of MBS) as we speak."
The timing matters when reading the September data. Lawler notes that commitments made in the second half of September would almost certainly not settle until October or later, so they would not appear in the September reports. The September rise therefore suggests the companies had already picked up their buying a bit before Pulte's post. Any larger effect from what he described would show up in later reports.
Two different balance sheets
The bond holdings are only part of each company's mortgage portfolio, and the two companies are moving in different directions.
Change, end of 2025 to end of September
Fannie Mae
Freddie Mac
Agency MBS holdings
+$38.203B
+$12.46B
Total mortgage investment portfolio
+$47.767B
−$3.792B
Corporate and operating liquidity
+$14.413B
−$23.356B
Debt outstanding
+$41.971B
−$38.964B
Fannie has grown both its assets and its debt. Freddie has added mortgage bonds while shrinking its total mortgage portfolio, its liquidity holdings and its debt. By simple subtraction, Freddie's other mortgage assets fell about $16.3 billion this year, which more than offset its MBS gain. The same pattern appeared in September, when Freddie's total portfolio fell $3.012 billion even as its MBS holdings rose. Fannie's total portfolio rose $7.623 billion that month. Lawler does not explain why the two companies diverge.
Interest-rate risk remains high
Holding mortgage bonds funded with debt exposes a company to interest-rate risk even when credit losses are guaranteed. One standard way to measure that is to estimate how much market value would be lost if rates jumped suddenly. Both companies report sensitivity to an upward shock of 50 5, or half a percentage point.
Lawler says those measures were little changed in September and remain historically elevated. The specific readings appear in a table in the original post, not in its text. He doesn't discuss what is driving them or whether they concern him. The headline point is simply that the risk has not come down.
Yields fell, but spreads did not narrow
The 6 yield, a benchmark for bonds backed by newly issued mortgages, fell 34 basis points from the end of September to the end of October. That reversed exactly half of the 68-point rise from the end of August to the end of September, leaving the yield 34 points above its August level. Most intermediate- and long-term Treasury yields also fell in October.
Two details of that move stood out to Lawler. First, yields dropped faster right after September's readings on the Consumer Price Index and Producer Price Index, the main gauges of consumer and business-level inflation, came in somewhat hotter than expected. Hot inflation data would normally push yields up.
Second, the summer's surprisingly high correlation between interest rates and oil prices completely broke down in October. A high correlation between rates and tech stocks replaced it. Several analysts read this as investors rotating from tech stocks into bonds: selling tech pushes those shares down, and buying bonds pushes yields down, so the two fall together. Lawler's takeaway is that such correlations can be quite fleeting.
The CCMBS spread over Treasuries rose modestly in October after widening more sharply in September, though it ended the month below its October 16 peak. Lawler partly attributes the wider spreads since August to higher 7, the size of future rate swings priced into options.
The mechanism is prepayment risk. Homeowners tend to refinance when rates fall, returning bondholders' money just when reinvestment yields are low, and they stay put when rates rise, leaving bondholders stuck with low-yielding bonds. Larger expected rate swings make that risk more costly, so investors demand more yield.
News of the increased September purchases, Lawler adds, has so far had little market impact.
Municipal bonds partly recover
Municipal bonds had what Lawler calls a horrific September. Yields on top-rated AAA munis surged, and by significantly more than Treasury yields. The 8, a muni's yield divided by a comparable Treasury's, "exploded upward." For illustration, a 3.5% muni against a 4% Treasury gives a ratio of 87.5%. A higher ratio suggests munis are cheaper relative to Treasuries.
The trigger was a huge spike in selling by individual investors, including 9: selling a bond at a loss to offset capital gains elsewhere. Swings in munis' relative performance have often coincided with periods of sharply rising rates in recent years, Lawler says.
As taxable rates drifted lower in the second half of October, muni yields fell by significantly more than Treasury yields. Ratios came down from their September highs, but stayed above August levels and well above those from earlier in the year.
What remains open
The main unanswered question is whether Pulte's "large quantities" appear in the October and later reports. Lawler doesn't forecast that, or say how much buying would be needed to narrow spreads. The September evidence shows only a modest pickup, and one that markets have largely ignored.
Remember
Check promises against settled data: purchases committed in late September would not appear in September's reports.
Lower yields don't guarantee cheaper mortgages relative to Treasuries; volatility helped keep the MBS spread wider despite October's yield drop.
Market correlations can flip quickly: rates tracked oil over the summer, then tech stocks in October.
