That gap is the whole story, or at least the part worth sitting with. A national survey of virtual asset businesses, covering the first six months of the year and released at the start of October, put the slump in plain figures. Average daily trading volume dropped 44%. Won deposits fell 35%. The market value of crypto held on domestic platforms slid 33%. Sales at the exchanges themselves were down 41%. Profit did not drift lower. It collapsed, from KRW 374.8 billion in the prior half-year to KRW 81.6 billion.
Perhaps the most interesting aspect is how orderly the decline looks on paper and how uneven it feels once you split the market. Won-based venues still hold almost everything. Coin-only shops are a rounding error. Thin listings got thinner. Transfers out of the country slowed, but the pipes that remain are mostly whitelisted wallets, not small retail hops. Staffing barely changed. Losses in custody got worse. And a tax that has been promised for years is still scheduled for the first day of 2027.
What The Profit Collapse Actually Measures
Operating profit is not a vibe. It is what is left after the cost of running the shop. For a crypto exchange, that shop lives on spreads, fees, and the simple fact that people keep pressing buy and sell. When they stop, the cost base does not shrink at the same speed. Rent, compliance teams, matching engines, customer support, banking rails: those bills arrive whether Tuesday was busy or not.
The survey covered 26 registered virtual asset service providers from 1 January through 30 June. Seventeen were exchanges. Nine were custody and wallet businesses. The figures came from company submissions. They are not dressed up as official national statistics, which is worth remembering if you like your data with a footnote. Still, a half-year snapshot of the licensed market is about as close as outsiders get to the cash register.
I have found that percentage drops mislead when the starting point is a boom. The second half of 2025 was not a quiet baseline. KRW 374.8 billion of operating profit is a fat number for a domestic crypto industry. Falling to KRW 81.6 billion is still a business, just not the business operators had staffed for. A 78% decline can mean “the model broke” or “the cycle turned and the fixed costs did not.” Both readings are available. I lean toward the second, with a caveat about listings that never had real depth.
The Volume Math Behind The Miss
Average daily crypto trading volume nationwide fell to KRW 3.1 trillion from KRW 5.4 trillion. That is the 44% drop. Exchange sales fell 41% over the same comparison, which tracks closely enough that you do not need a conspiracy to explain the profit line. Fee income follows tickets. Fewer tickets, less income, and operating profit falls faster than sales because costs are sticky.
Think of a restaurant that lived on Friday nights. Cut Friday covers by nearly half and the chef still gets paid. The walk-in fridge still hums. Profit does not fall 44%. It falls more. Crypto venues are not restaurants, but the shape of the problem is familiar. High operating leverage feels brilliant on the way up and rude on the way down.
Volume is the oxygen. Profit is what is left after the building keeps breathing.
A separate look at the first half put combined turnover on the five largest domestic venues at $366.58 billion, down 54.6% from a year earlier. The largest venue took a bigger share of a smaller pie. Market share gains in a shrinking pool are a strange kind of victory. You can win the room and still earn less. Reports around that period also said the leading venue’s first-half net profit fell about 74%, while the second-largest swung from profit to a net loss. Operating profit for the sector and net profit for two giants are not the same line item, but they rhyme.
Deposits Left, Accounts Stayed
Won-denominated deposits available for trading fell 35%, to KRW 5.2 trillion from KRW 8.1 trillion. The value of crypto held through Korean exchanges dropped 33%, to KRW 58.9 trillion from KRW 87.2 trillion at the end of 2025. Cash on the sidelines and coins on the books both shrank. That is not a story about one coin having a bad month. It is a story about the domestic pool getting smaller in both currencies.
Accounts did the opposite, barely. Tradable accounts stood at 11.175 million at the end of June, up 0.4% from 11.126 million six months earlier. The most common user age group shifted from people in their 30s to people in their 40s. Accounts holding less than KRW 1 million in virtual assets rose by 370,000, to 8.63 million.
Read that twice. The typical registered account got smaller, older, and less active as a source of fees. Registration is not engagement. A login that never places an order is a row in a database, not a customer. In my experience, industries love to quote the account number when the revenue number is embarrassing. Here the survey is honest enough to show both.
- Operating profit: KRW 81.6 billion, down 78% from KRW 374.8 billion
- Average daily volume: KRW 3.1 trillion, down 44% from KRW 5.4 trillion
- Exchange sales: down 41% versus the prior half
- Won deposits: KRW 5.2 trillion, down 35% from KRW 8.1 trillion
- Domestic crypto value on exchanges: KRW 58.9 trillion, down 33% from KRW 87.2 trillion
- Tradable accounts: 11.175 million, up 0.4%
- Small accounts under KRW 1 million: 8.63 million, up 370,000
The small-account surge is easy to romanticize as “retail is still here.” Maybe. It is also what a market looks like when larger balances have already left or marked down, and what remains is a long tail of modest holdings. Fee schedules built for active traders do not harvest much from a wallet under a million won.
A Price Effect Hiding Inside The Holdings Number
KRW 58.9 trillion is a mark-to-market figure, not a count of coins. If prices fall, the same stack of tokens is worth less. If people sell and withdraw, the stack itself shrinks. The survey does not hand you a clean split between those two forces. Earlier central-bank figures, cited in the spring, showed domestic crypto holdings falling by more than half from KRW 121.8 trillion in January 2025 to KRW 60.6 trillion by February 2026. Part of that slide was tied to stronger demand for local equities. So the first-half exchange survey sits inside a longer drawdown, not a sudden cliff that appeared on 1 January.
That longer arc matters. A 33% drop from end-2025 to end-June is severe. A halving from the January 2025 peak is the wider weather. Operators who hired into the peak were staffing a market that, by early 2026, had already given back a huge share of notional value. Profit in the first half is the bill for that mismatch arriving late.
Won Venues Still Own The Room
Concentration did not loosen while profits sank. Won-based exchanges held KRW 58.5 trillion of the KRW 58.9 trillion in domestic crypto value at the end of June. Coin-only exchanges held KRW 330 billion, roughly 0.6%. Trading was even more lopsided. Won-market venues generated about KRW 3.1 trillion in average daily volume. Coin-only venues recorded KRW 380 million. Volume fell 44% on won platforms and 55% on coin-only venues versus the previous half.
If you trade only coin pairs, you are already a specialist. In a risk-off stretch, specialists leave first. The fiat on-ramp is where salaries, bonuses, and stock-sale proceeds actually arrive. Korea’s crypto market has long been a won market with a crypto coating. The first half of 2026 did not change that architecture. It made the coating thinner.
A new turnover measure in the survey put monthly trading turnover between 100% and 201% for won-based exchanges. Coin-only platforms sat between 2% and 9%. For scale, the same review placed KOSPI monthly turnover at 25% and KOSDAQ at 43% during the first half. Crypto, even in a slump, still churns harder than the stock market. That is not automatically healthy. Turnover above 100% a month means the average unit of inventory is theoretically traded more than once. It can mean deep liquidity. It can also mean a smaller set of coins being passed around by the same active cohort while everyone else watches.
| Slice of the market | End-June snapshot | What changed |
| Won-based exchange holdings | KRW 58.5 trillion | Still nearly the entire domestic book |
| Coin-only holdings | KRW 330 billion | About 0.6% of domestic value |
| Won-market daily volume | About KRW 3.1 trillion | Down 44% versus prior half |
| Coin-only daily volume | KRW 380 million | Down 55% versus prior half |
| Monthly turnover, won venues | 100% to 201% | Far above local equity turnover |
| Monthly turnover, coin-only | 2% to 9% | Thin by any standard |
| KOSPI monthly turnover | 25% | Benchmark, not a crypto figure |
| KOSDAQ monthly turnover | 43% | Still well below won-crypto churn |
I keep circling that equity comparison because it explains the summer headlines better than any slogan about “adoption.” In late July, average daily trading across the five main venues had fallen to KRW 597.8 billion through the prior Tuesday. That was equal to 1.59% of the KRW 37.6 trillion average daily turnover on the KOSPI. Crypto did not vanish. It became a side pocket next to a stock market that Koreans were actually using.
Did The Money Simply Rotate Into Stocks?
Some of it, yes. Not all of it, and not in a straight line. The holdings decline from early 2025 into February 2026 lined up with stronger demand for domestic shares. That is a rotation story, and rotation stories are popular because they feel tidy. Sell the thing that stopped going up. Buy the thing your group chat is posting about. Korea’s retail culture is intense in both markets, so the channel exists.
Tidy is not the same as complete. Prices falling reduce the won value of coins that never left the exchange. Some balances moved to self-custody. Some moved abroad, though external transfers also fell, which cuts against a simple “everyone wired it overseas” tale. And a slice of users did nothing. They kept the account, shrank the balance, and waited. The age shift toward the 40s fits a wait-and-see cohort better than a stampede of new speculators.
August then ruined the one-way narrative, which is what markets do when you get too confident. During a bitcoin rally that month, daily volume on the largest venue jumped 273% to $1.84 billion. The second-largest rose 132.9% to roughly $934.9 million. Same industry, same year, opposite month. The first half was a drought. August was a squall. Profit reported for January through June cannot include that squall, which is why half-year headlines age fast.
If you only read the 78% figure, you miss the reflex. Korean crypto volume still answers when the global tape turns. It just answers from a lower base, with less cash parked in won deposits, and with a stock market next door that has been louder. The reflex is not dead. The wallet funding it is lighter.
Listings That Only One Shop Will Touch
Liquidity worries showed up most clearly in assets listed on a single Korean platform. Unique crypto assets in circulation fell 5% to 673. Single-exchange listings declined to 234 from 296. Those exclusive names represented KRW 600 billion, about 1% of domestic crypto value. Ninety-three of them, equal to 40%, each had market values of KRW 100 million or less.
KRW 100 million is not a market. It is a group chat with a price. The survey warned, correctly, that these names are exposed to low liquidity and sharp moves. A 1% slice of value does not move the national profit number. It does move the reputation of a venue that listed them, and it moves the loss account of whoever is stuck as the bid. In a bull phase, exclusive listings are a marketing feature. In a 44% volume drawdown, they are inventory nobody wants to warehouse.
Delisting is slow, political, and full of complaints. Leaving a dead market up is cheaper until it is not. I suspect the drop from 296 exclusive listings to 234 is the start of a cleanup, not the end. Venues protecting a thinner fee pool have less reason to carry tokens that generate noise and almost no turnover. The coin-only turnover range of 2% to 9% a month already says a lot of listed inventory is ornamental.
Listing health, end of June: Unique assets in circulation: 673 (down 5%) Single-exchange listings: 234 (from 296) Value of exclusive listings: KRW 600 billion (about 1%) Exclusive names at or under KRW 100 million: 93 (40%)
Transfers Slowed, And Most Of Them Were Already Known
Activity off the order book weakened too. External transfers from Korean exchanges dropped 41%, to KRW 62.8 trillion from KRW 107.3 trillion in the previous six months. Transfers covered by the travel rule totaled KRW 9.5 trillion, or 15%. Whitelisted transfers to approved overseas entities and personal wallets totaled KRW 51.7 trillion, or 83%. Transfers under KRW 1 million made up KRW 1.6 trillion, only 2% of value, but involved 63% of users who moved funds externally.
That split is the cleanest portrait of two publics sharing one pipe. A small number of larger, pre-approved flows carry almost all the won. A large number of people move pocket change. The travel-rule slice, at 15%, is the regulated handshake between institutions. The whitelist slice is the established path for people who already did the paperwork. Neither looks like a panic exit. Both look like a smaller version of the previous habit.
A 41% drop in external transfers alongside a 33% drop in domestic holdings and a 35% drop in won deposits suggests the system cooled rather than ruptured. Money did not stampede out the side door in greater proportion. The side door itself got less use. If anything, the whitelist dominance says the remaining cross-border flow is identified, not shadowy. That will matter when tax authorities start asking exchanges for gain calculations.
Custody Got Quieter And Lost More Money
Custody and wallet businesses had their own version of the same weather. Customer accounts rose 2% to 792. Assets in custody fell 25% to KRW 230.4 billion. Operating results deteriorated from a KRW 9.3 billion loss in the second half of 2025 to a KRW 18.6 billion loss in the first half of 2026.
Seven hundred ninety-two accounts is a boutique book, not a retail crowd. These are the businesses holding assets for clients who do not want to leave coins on an exchange matching engine. The account count edged up while the assets fell by a quarter and the loss doubled. That is a painful combination. You added relationships and subtracted collateral. Revenue in custody often scales with assets under watch. Costs scale with licenses, audits, and people who understand key management. Shrink the assets, keep the people, and the loss widens. Exactly what the half showed.
Exchange staffing, by contrast, was almost flat. The sector employed 2,021 people at the end of June, down by 10 from six months earlier. Anti-money-laundering staffing edged up by one person, to 213. Nobody gutted the compliance desk to save the quarter. That is either discipline or inertia. Given how hard licensed status is to win in this market, I would bet discipline. You do not fire the person who keeps the license alive because February was slow.
Headcount held. The profit did not. That is what a regulated downturn looks like when the license is worth more than the quarter.
Reading the staffing line against the profit line
Why Profit Fell Faster Than Volume
A 44% volume drop and a 78% profit drop are not a contradiction. They are a lesson in cost structure. Start with fee rates. In competitive won markets, headline fees get discounted for active tiers, market makers, and promotional periods. When retail goes quiet, the remaining flow is often the discounted flow. Blended take-rate falls even if the sticker price does not. Sales down 41% on volume down 44% is consistent with a slightly worse mix, not a fee war, but it is not a richer mix either.
Then add the things that do not flex. Banking partnerships. Real-time monitoring. Insurance. Travel-rule vendors. Office space in Seoul that was signed when the second half of 2025 looked permanent. A ten-person reduction on a base of 2,021 is a rounding error against a profit hole of nearly KRW 300 billion. Labor was not the lever they pulled. They may not have had a lever that pulls that fast without touching the license.
There is also the inventory and incentive layer. Campaigns, listing support, and liquidity programs are easier to justify when daily volume is KRW 5.4 trillion. At KRW 3.1 trillion they look like subsidies. Cut them and you protect profit, but you also risk the turnover that still separates you from a dead order book. Keep them and the operating line suffers. The first-half result looks like a compromise: listings trimmed at the fringe, core won books kept open, profit allowed to fall.
- Volume fell hard, so gross fee opportunity fell with it.
- Sales fell almost in line, so pricing did not save the top line.
- Fixed and semi-fixed costs barely moved, so the drop hit profit twice as hard.
- Custody losses widened, adding a second drain beside exchange operations.
- August’s volume spike arrived too late to repair the half that had already closed.
The User Who Registered And Then Did Very Little
Eleven million tradable accounts is a large number in a country of about 52 million people. It is also a stale number if most of those accounts hold under KRW 1 million and do not trade. The rise of 370,000 small accounts can mean new sign-ups who never funded, old accounts that sold down, or both. The survey does not separate those cases. Either way, the economic customer is not the registered customer.
The shift in the most common age band, from the 30s to the 40s, is a quiet tell. Younger cohorts often dominate the manic phase of a retail market. A move toward the 40s can mean the speculative edge aged, or that younger users left for equities and the remaining book looks older. It can also mean nothing dramatic. Age bands move when a few hundred thousand accounts change classification. I would not build a generation theory on a 0.4% account increase. I would build a caution on it. The market did not lose its user file. It lost its urgency.
Urgency is what fee businesses sell. A person who checks prices twice a week and holds a small bag is not the same unit of demand as a person who turns over their balance several times a month. The turnover figures say the active core is still hyperactive relative to stocks. The deposit figures say that core is funding itself with less cash. Hyperactive on a smaller float produces drama and less profit. That is the first half in one sentence.
What A 22 Percent Tax Does To A Thinner Market
Separate from the survey, traders are scheduled to face a tax change on 1 January 2027. The plan that has been maintained is a 22% tax on annual crypto gains above KRW 2.5 million, with authorities preparing implementation guidance for domestic exchanges. Two and a half million won is not a high threshold. At recent exchange rates it is a modest profit, the kind a small account might book in a single lively month and then give back.
Will that tax explain the first-half slump? No. It was not in force. Can it shape behavior in the second half of 2026 and into the start of 2027? Yes, if people believe the date will hold. Korea has postponed this levy before. A postponed tax teaches two lessons at once: the bill is real, and the calendar is negotiable. Some holders will realize gains before the rule bites. Some will do nothing and hope for another delay. Exchanges will spend money they do not have, relatively speaking, on reporting systems.
That last point is easy to skip. Implementation guidance means product work, legal work, and customer education. Those are costs. They land on the same operators who just posted a 78% profit decline and who did not cut staff in any meaningful way. A tax that is collected at the platform is also a reason to keep balances on a reporting venue rather than a coin-only shop with no fiat rail. In that narrow sense the levy reinforces the won-exchange oligopoly. It does not refill the KRW 8.1 trillion deposit pool.
I would watch the threshold more than the rate. Twenty-two percent is noticeable. KRW 2.5 million is the tripwire that pulls ordinary accounts into the system. If small accounts are already the majority of the file, a low exemption is a mass-market compliance event, not a whale event. The transfer data already shows that most people who move money externally move little of it. Tax reporting will meet that same long tail.
How This Compares With A Normal Risk-Off Stretch
Crypto profit is cyclical everywhere. Korea is not unique in watching fees collapse when retail steps back. What is local is the degree of fiat concentration, the intensity of the equity alternative, and the regulatory perimeter around who may offer won pairs. When the stock market is printing KRW 37.6 trillion of average daily turnover and crypto is printing KRW 0.6 trillion on a dull July stretch, the opportunity cost is not theoretical. It is on the same phone.
Global rallies still leak in. August proved that. A 273% daily volume jump does not happen in a market that has structurally exited. It happens in a market that is under-positioned and quick to chase. Under-positioned is exactly what a 35% drop in won deposits describes. Chasing is what 100% to 201% monthly turnover describes. The combination is jumpy, not abandoned.
Operators will talk about diversification: staking, custody, tokenized funds, brokerage of overseas products if the rules allow. The custody line in this survey is a warning label on that pitch. Accounts up, assets down, losses wider. New business lines do not automatically replace spot fees. They can add losses while the old engine idles. Anyone modeling a recovery should separate “volume comes back” from “we built a second profit center.” The first is cyclical and plausible. The second is unproven in these numbers.
Reading The Survey Without Fooling Yourself
A few limits belong next to the headlines. Company-submitted figures can be consistent and still not be audited national accounts. Operating profit is not cash flow and not net profit. The August spike sits outside the window. Price moves contaminate the holdings total. Account counts include people who are effectively dormant. Exclusive-listing counts include tokens that should not be called markets.
Even with those limits, the direction is hard to argue with. Sales, volume, deposits, holdings, external transfers, and custody assets all fell together. Accounts and compliance headcount did not. That pattern is a demand shock inside a regulated shell, not a run on the shell. If it were a trust crisis, I would expect transfers to spike, not drop 41%, and I would expect the whitelist share to look less orderly.
The political reading is available too, and I would keep it modest. A profit collapse makes the industry a smaller taxpayer-in-waiting and a louder lobby for delay. It does not, by itself, prove the rules caused the slump. Volume fell globally in stretches of 2026 whenever bitcoin cooled. Korea amplified that with an equity bull run next door. Blaming only the regulator, or only the chart, throws away half the evidence.
What Would Have To Change For Profits To Heal
Healing is not mysterious. Deposits have to refill, or prices have to rise enough that existing inventory throws off more notional turnover, or both. Fee rates would need to hold. Costs would need to stop growing. August showed the volume switch still works. It did not show that the deposit switch has flipped back to the KRW 8 trillion area. Without that cash, rallies are shorter, because the same won is being recycled rather than new savings arriving from paychecks and stock profits.
A rotation back from equities is the local catalyst people will hunt for. It requires the stock trade to bore them, or to lose them money, and crypto to offer a cleaner story at the same time. Those conditions arrive together less often than commentators imply. They can arrive. The January 2025 holdings peak of KRW 121.8 trillion is the memory the industry is measuring itself against. Getting halfway back would already transform the fee pool. Nothing in the June survey says that path is open this quarter.
Listing quality is the unglamorous lever. Cutting exclusive names that trade at pocket-change market caps will not restore KRW 300 billion of operating profit. It will reduce complaint risk and the chance of a messy print on a dead book. In a market this concentrated, reputation at the two or three won venues is the franchise. The fringe tokens are not.
Rough healing checklist:
Deposits closer to KRW 8 trillion than KRW 5 trillion
Daily volume holding above the KRW 3 trillion area for more than a rally week
Blended fees stable, not discounted away
Custody assets growing, not just custody accounts
Tax systems built before 2027, not during a volume spikeThe Concentration Risk Nobody Is Paid To Mention
When 99% of domestic crypto value sits on won-based venues, a profit slump is not evenly shared. Coin-only shops were already a 0.6% sideshow, and their volume fell faster. The pain that moves policy, banking access, and public perception sits with the fiat platforms. That concentration is efficient in a boom. In a bust it means a handful of income statements are the industry. The sector profit figure of KRW 81.6 billion is, in practice, their figure.
There is a user-side version of the same risk. If your exchange is your market, a quiet book is your quiet book. You cannot hop to a deep coin-only alternative inside the country and find KRW 3 trillion of daily flow. It is not there. Overseas venues exist, and the whitelist data says some people use them, but the domestic price people screenshot is still the won price. Concentration makes the downturn feel national even when the cause is partly global.
I do not think that structure breaks in the next year. Banking rails and real-name rules are hard to clone. The tax guidance will likely be written for the venues that already file. Smaller players get a compliance bill without the volume to pay it. The first-half loss at custody businesses is a preview of what “be a regulated crypto company” costs when assets are not growing.
A Note On The Summer That Followed The Survey
The survey window closed on 30 June. July, at least the stretch reported late that month, was worse on a daily-volume basis than the half-year average. KRW 597.8 billion a day is not KRW 3.1 trillion. If July had been the whole story, the 78% profit drop would have looked like a midpoint, not a floor. Then August arrived with a bitcoin move and daily prints back in the billions of dollars on the top venues.
Hold both months in your head. A market that can fall to well under KRW 1 trillion of daily flow and then triple on a rally is not stable, and it is not finished. Profit will gap around those swings. Analysts who annualize June will be wrong. Analysts who annualize August will also be wrong. The honest range is wide, and the deposit number is the anchor. Cash in the account is harder to fake than a single green day.
That is why I keep returning to KRW 5.2 trillion. Volume can be borrowed from leverage and from the same coins turning over. Deposits are closer to intent. A 35% drop in intent, sitting beside an 0.4% rise in permission to trade, is the behavioral core of this report. People kept the key. They took a lot of the money off the table, or the table marked the money down, or both.
What Traders And Operators Can Actually Use
If you trade the won books, the practical read is liquidity quality, not the national profit headline. The books that matter are still the won books. Turnover remains high versus equities, so spreads on majors can stay reasonable even in a quieter half. The danger is the exclusive listing, the 40% of single-venue names with tiny market value. Those are where a quiet month becomes a hole in the bid. Treating them as investments because they have a Korean ticker is how small accounts become smaller accounts.
If you run a venue, the practical read is operating leverage and the 2027 build. Ten fewer employees will not restore the first half. A clearer listing standard might protect the brand. A custody product that loses more as assets fall is not a hedge against spot. The cost of tax reporting is coming whether volume returns in September or in 2027. Building it during a slump is miserable. Building it during the next August is impossible, because every engineer will be on incident calls.
If you allocate outside Korea and only watch this market as a sentiment gauge, use it as a retail-intensity gauge, not a bitcoin oracle. Korean volume spikes when global price moves and local cash is available. The first half said local cash was less available. The equity turnover comparison said attention had a competing home. Neither cancels the August reflex. Together they say the next spike needs a reason, and the reason has to beat a stock market that has been taking the oxygen.
The Number That Should Have Been The Headline
The 78% is the number that travels, and it should. It is accurate to the survey, dramatic, and easy to repeat. The number I would tape above it is the account figure next to the deposit figure. Permission up. Cash down. That is how a fee business starves without a customer revolt. No queue outside the office. No mass closure of accounts. Just a thinner float, an older modal user, and a long list of balances too small to matter to the income statement.
Add the listing fringe, the custody loss, the flat headcount, and the tax date, and you have a market that is regulated, concentrated, and waiting. Waiting is not a strategy for an exchange. It is a description of the user. The exchange still has to pay for the wait. That, more than any single red candle, is what drove operating profit from KRW 374.8 billion to KRW 81.6 billion.
Will the second half look like June, like July, or like August? The survey cannot say, and anyone who pretends otherwise is selling a chart. What it can say is that the domestic machine entered the summer with less fuel, the same number of licensed engines, and a compliance team that did not shrink. Fuel is deposits and attention. Engines are the won venues. The rest is noise at the edge of the order book, where KRW 100 million pretends to be a market.
I will be watching the deposit line more than the profit line from here. Profit will follow if the cash comes back and stays long enough to survive a quiet month. If it does not, the next survey will rhyme with this one, staff counts and all, and the 78% will stop looking like a shock. It will look like the new run rate, with Augusts as interruptions. That is a colder outcome than a crash. It is also, on these figures, the one already underway.
]]>
