A founder forges the crown alone, and most of the weight of starting a business comes before it is worn.
JIAHON Insights | Marcus, Senior Vice President | 31 August 2026 | Data: Small and Medium Enterprise and Startup Administration (Ministry of Economic Affairs), Ministry of Finance, Directorate-General of Budget, Accounting and Statistics (Executive Yuan), Taiwan Lottery
The metaphor needs pinning down at the outset, or this becomes a motivational piece.
Wearing the crown means becoming the person in charge of a company. From that day, the founder is the company’s sole signature to the outside world, the sole decision-maker inside, and the person who signs on the bank’s credit form as joint and several personal surety. The founder is the company’s king.
The Chinese saying “who wears the crown must bear its weight” is usually read as: once the crown is on, the responsibility is heavy.
But look at the numbers and the weight is not there. The weight comes before the crown goes on.
Nobody else crowns the founder; the founder spends ten years forging the crown. And forging it costs far more than wearing it.
This article sets out to put a number on that forging cost.
(Survival rates, recovery rates and expected values quoted here come from another article, Choosing a road to NT$100 million.)
1. The entry ticket is what keeps a founder alive through the first three years
Start with a set of numbers.
The actual survival curve of new businesses in Taiwan: 24.0% are gone in the first year, 67.1% remain at year three, and 44.0% at year ten.
The first-year cut is more than four times heavier than any of the following nine years. The average failure comes at year 3.5.
In other words, the vast majority of business deaths are concentrated in the first three or four years. Past that stretch, the annual death rate drops to between 2% and 4%, no different from ordinary businesses.
So the real question is not “how to grow big” but “how not to die in the first three years”.
The set of things that keeps a person alive through those years is what this article calls the entry ticket.
It is not resources, not capital, and not “having ideas”. It is a very concrete set of things that can be checked one by one.
2. Four mechanisms, all of them lowering the death rate
The ticket works through four mechanisms, each matching a typical way to die in the first three years:
| How a founder without a ticket dies | Why a founder with a ticket survives this |
|---|---|
| Does not know why the customer is stuck on a specification; three revisions miss; the money runs out | Has worked that process; gets it right the first time |
| Suppliers demand cash; working capital cannot keep up | Track record gets the supplier to grant 90-day terms |
| Cold calls for a year without getting onto the customer’s approved vendor list | Knows the decision-makers; qualification in three months |
| Banks lend against collateral alone; cash flow seizes | Experience gets the bank to approve a credit line |
All four lower the death rate; none raises the ceiling.
The ticket does not make a founder earn more than others. It keeps the founder alive longer, long enough to earn.
In numbers: ten-year survival is 44.0% for all new businesses, and estimated at around 65% for founders with this set of things. What the ticket buys is the area between those two lines.
The second layer: which rung of the gross margin ladder a founder can stand on
Once survival is secured, a second effect comes into play, and this one can be quantified.
Running a regression across 1,283 tax industry subcategories in the Ministry of Finance’s 2025 (ROC year 114) industry profit standards gives a very stable relationship between gross margin and net margin: net margin = 0.348 × gross margin + 1.51 (r = 0.897). In other words, a business’s profitability is decided by its gross margin, and gross margin is decided by which problems the founder can solve.
| Gross margin | Net margin | Odds of NT$100 million in ten years, as a multiple of lottery odds | Ticket this rung requires |
|---|---|---|---|
| 19% | 8.1% | 21× | No ticket; trading spread and turnover |
| 30% | 11.9% | 33× | Knowing which items sell and having suppliers willing to supply |
| 40% | 15.4% | 45× | Reading specifications and choosing the right item for the customer |
| 50% | 18.9% | 58× | Knowing why the customer’s process has problems |
| 57% | 21.3% | 67× | Solving problems others cannot, plus exclusive agency rights |
| 65% | 24.1% | 77× | The customer writes the company into its specification |
(19% is the official standard’s typical figure for wholesale; 57% is an industry estimate for niche technical distributors.)
The multiple runs from 21 to 77, a factor of 3.7. Each rung up requires a deeper layer of ticket.
Anyone can enter the 19% rung, so nobody makes money there. The 65% rung requires the customer to write the company into its specification, which is not won in a negotiation but accumulated over ten years.
Gross margin is not a price the market sets; it is the exchange rate of the ticket.
But the honest version has to be finished
Climbing the ladder does not make founding a business a good bet. Run through the same model, the 20-year net expected value is −NT$16.02 million at 19% gross margin and −NT$15.57 million at 65%: the multiple differs by a factor of 3.7, but the expected value differs by NT$450,000.
That is because the expected value is dominated by the 43% who survive without reaching NT$100 million; the NT$100 million outcome carries too little weight.
The ticket’s first layer keeps the founder alive. Its second layer decides how high the founder stands while alive.
But it changes the thickness of the tail, not the expected value.
That is a cold sentence, but an honest one. The ticket is no promise of wealth. It turns a losing bet into a losing bet that is worth more: what the founder earns, builds and learns along the way is more than in the version without a ticket.
How to tell whether one has it
The ticket is hard to self-assess, because everyone thinks they have good judgment. A more reliable test is these two lines:
Knowing why the customer makes that decision, and having people who will take the call.
The first half is judgment in concrete form: not “I have ideas” but “I know why that buyer is stuck on this specification, why they cut this budget, why they chose another supplier”. The second half is relationships.
Both can be checked: think about the last proposal that was rejected, and whether the real reason was known at the time; think about who would pick up the phone today.
3. Why one person wears it
Taiwan’s small and medium enterprises by form of organization:
| Form | Share | Number of people |
|---|---|---|
| Sole proprietorship and other | 57.47% | One person by definition |
| Limited company | 31.24% | Cannot be determined (single-member limited companies are legal) |
| Company limited by shares | 8.76% | Several shareholders |
| Partnership | 2.52% | Clearly several people |
A partnership is the formal structure under the Business Registration Act for “several people contributing capital and running a business together”; it is the legal name for several people starting out together. It accounts for 2.52%.
This share is not shrinking; it never grew in the first place.
For six in ten businesses, the form of organization already tells how many people are involved, and 96% of those are one person.
This is no accident; it can be calculated
The reason lies in the cost structure of the ticket: the ticket’s value does not scale with the number of people, but the opportunity cost does.
When one person with a ticket starts a business, they give up one salary. When three people, each with a ticket, start together, they give up three, and three high salaries, because people with tickets are already well paid.
Using a 20-year model (assuming that one more co-founder raises the owners’ annual income by 1.5 times and survival by another 5 percentage points, which is already optimistic):
| Co-founders | 20-year expected value | Baseline (each stays employed) | Net difference |
|---|---|---|---|
| 1 | NT$47.60 million | NT$42.00 million | +NT$5.60 million |
| 2 | NT$84.75 million | NT$84.00 million | +NT$0.75 million |
| 3 | NT$117.85 million | NT$126.00 million | −NT$8.15 million |
Each added person drops the expected value a rung. The moment a third person joins, the whole thing turns negative.
The reason is simple: opportunity cost rises linearly with the number of people (NT$42 million per person over 20 years), and the value of the business does not. Two people do not build twice the business, especially in distribution, where the scale of the business is set by agency rights and customers, not by the number of founders.
A different angle makes it clearer
Suppose the business needs someone who understands the process and can talk to customers. There are two choices:
Make them a co-founder: they give up an annual salary of NT$2 million, an opportunity cost of NT$20 million over ten years, all staked on the same bet; if the company fails, both go to zero.
Hire them as a senior employee: NT$1.5 million a year, NT$15 million over ten years. That is an expense, not a stake. They are paid every month and carry no downside risk.
A co-founder costs NT$5 million more than an employee and ties two people’s risk together.
That is why the typical structure of a Taiwanese SME is one owner carrying the ticket plus a team of employed staff. It is not that the owner wants to keep everything; the arithmetic of the structure allows no more than one person to carry it.
The Silicon Valley model can have co-founder teams because venture capital buys that opportunity cost with equity; someone pays it up front. Taiwanese SMEs have no such mechanism and carry it themselves.
4. When to put it on: where four constraints meet
The actual age of entry is not random. Among SME employers, 27.72% are under 45, and the mode falls between 45 and 59.
In other words, most people put on the crown after forty.
This is not coincidence; it is where four constraints meet:
- Accumulation takes ten years. Getting from the bottom to a position near the decisions (knowing why customers decide as they do, and knowing the people who decide) takes about ten years. Before thirty, a complete ticket is rare.
- Tickets depreciate. Industry relationships decay: buyers move, engineers retire, processes change generation. Contacts usable at forty are less than half usable at fifty-five. Enter too late and the ticket is already losing value.
- The operating period has a minimum. Running a full accumulation curve takes ten years. Enter at forty, take stock at fifty, and ten years remain to adjust or exit. Enter at forty-five and there is one chance.
- The way back has a window. Fail at forty and a return to employment is possible. Fail at fifty and the re-employment discount falls from 70% of the previous pay to below 50%, and that discount directly sets the expected value of the failure branch.
The four lines meet between thirty-eight and forty-two.
So forty is not a round number; it is where the four constraints meet. The white paper’s employer age distribution is the measured result of that intersection.
One more signal: the share of SME employers aged 55 and over rose from 37.85% in 2023 to 39.59% in 2024, and the number of own-account workers has fallen for five consecutive years since 2020. The people wearing crowns are getting older, and no one is stepping up behind them.
5. Can the crown be handed over?
This section is for those already wearing it.
The ticket has an uncomfortable property: it is attached to the person, not to the company.
Taiwan’s institutions acknowledge this everywhere: bank credit requires the owner to stand as joint and several personal surety (the bank lends to the person, not the company); agency rights are negotiated with a person, and a change of owner usually means renegotiation; customer qualifications follow the person; and when a company is sold, the buyer asks the owner to stay on under contract.
That gives an uncomfortable equation:
Company valuation = transferable assets (high recovery) + the owner’s ticket (zero recovery)
The larger the ticket’s share in a business, the more profitable it is, and the harder it is to sell.
Put another way: the more the owner is worth, the less the company is worth.
This is also the deeper reason liquidation recovery in distribution is around 20%: a buyer can acquire inventory, receivables and the customer list, but not the fact that customers take the owner’s call.
Five kinds of ticket, and how far each can be moved
| Kind | Content | Can it be moved? |
|---|---|---|
| Knowledge | Which material solves which process problem, how specifications match | Yes |
| Handling | How customers ask, how to answer, when to step back | Yes |
| Relationships | Who will take the call | Partly |
| Credit | The owner’s personal surety on bank credit | No (required by law) |
| Direction | Where to go next, whether to take on this line | No |
The first two can be codified, written into documents and handed to systems. The third can be partly transferred by deliberately not letting relationships attach to oneself, for example by having others meet customers in a technical role rather than “the boss coming in person” every time.
The fourth and fifth cannot be moved, and those two happen to be the valuable ones.
A tension that has to be admitted
What can be codified is also what competitors can copy easily. What cannot be codified forms the moat, and that is exactly what cannot be moved onto the company.
Transferability and competitive advantage are negatively correlated.
There is one exception: accumulated data assets, such as years of purchasing history, gross margin by item and customer buying cycles. Competitors cannot get them, and they work without the owner present. This is the one thing that raises both the moat and the recovery rate.
The test
Do not ask “would the company collapse without the owner?” That is a survival-style question with too low a bar; a company can run on inertia for a long time.
The right question is:
If the owner is away for three months and the company meets a problem it has never seen, does it reach the same answer the owner would?
If yes, the ticket has been moved. If not, but business carries on, that is inertia not yet used up.
6. The owner’s real job
Joining the five sections, the owner’s job comes down to three actions, in order:
One, accumulate it. Ten years, in a position near the decisions. There is no shortcut, and its cost is certain: forgone salary, a locked-in industry, irreversible path dependence.
Two, use it. Use it while the ticket is fresh: around forty, while relationships are intact, the way back is open and the operating period is long enough. That is where the four constraints meet, not a question of courage.
Three, move it. Move it from oneself onto the company. The first two kinds can be codified, the third partly transferred by depersonalizing customer contact, and the fourth and fifth cannot be moved; but knowing which cannot be moved is itself valuable, because it decides what the company is actually worth.
Most people do the first two and stop at the third, because doing it makes oneself look less important, which runs against instinct.
But that is where the difference lies: do the first two and there is a decent business. Do all three and there is a company.
Last word
Back to the saying.
The weight of the crown is not on the head; it is in the ten years spent forging it. And once it is on, the real work is not keeping it steady but making it stand without the wearer.
FAQ
Q1: Do nine in ten new businesses really fail in the first year?
Not nearly that many. On an operating basis, using the Ministry of Finance tax registry calibrated with the Industry and Commerce Census, 24.0% of new businesses in Taiwan are gone in the first year, 67.1% are still operating at year three, and 44.0% remain at year ten. The popular claim that “nine in ten fail in a year and 1% are left after five” does not match the tax registry data; before quoting any survival rate, ask which basis it is measured on. The period to guard against is the first three or four years: failure comes on average at year 3.5, and past that stretch the annual death rate is about 2% to 4% (see section 1).
Q2: What share of Taiwanese businesses are partnerships, and how does a partnership differ from a single-member limited company?
According to appendix table 11 of the SME White Paper, partnerships account for just 2.52% of SMEs, against 57.47% sole proprietorships, 31.24% limited companies and 8.76% companies limited by shares. A partnership is the form under the Business Registration Act in which several people contribute capital and run the business together. Under the Civil Code, when partnership assets cannot cover its debts, each partner is jointly and severally liable for the shortfall. Shareholders of a limited company are in principle liable up to their capital contribution, and a single person can form one. When an owner wants others on board, the common practice among Taiwanese SMEs is for one responsible person to carry the business and bring in the needed specialists as employees (see section 3).
Q3: What gross margin does a wholesale or distribution business need to make money?
Under the Ministry of Finance’s 2025 (ROC year 114) industry profit standards, the typical gross margin for wholesale is about 19%, which converts to a net margin of about 8.1%. The relationship this article draws from 1,283 industry subcategories is “net margin ≈ 0.348 × gross margin + 1.51”. A business can use its own gross margin to estimate a rough net margin range, then compare it with the actual books to see whether expenses are out of control. The relationship is an average across all industries; individual companies have different cost structures, so results will vary (see section 2).
Q4: Why does the owner have to give a personal surety when the company borrows from a bank?
Because the bank is assessing the owner as a person, not just the company’s financial statements, SME credit often requires the owner to sign as joint and several surety. Once that is signed, the company’s limited liability offers little protection for that loan: if the company cannot repay, the bank can pursue the owner personally. That is also why the “credit” kind of ticket cannot be transferred to the company or to a successor; when the owner changes, credit terms usually have to be renegotiated with the bank (see section 5).
Appendix: the grade of the data in this article
| Item | Value | Grade | Source |
|---|---|---|---|
| Ten-year survival (operating basis) | 44.0% | Hard | Ministry of Finance tax registry, calibrated with the Industry and Commerce Census |
| First-year death rate | 24.0% | Hard | Same |
| Average failure point | Year 3.5 | Hard | Same |
| Form of organization | Sole proprietorship 57.47% / partnership 2.52% | Hard | SME White Paper, appendix table 11 |
| SME employer age distribution | Under 45: 27.72% | Hard | SME White Paper, table 5-1-4 |
| Own-account worker trend | Down five consecutive years since 2020 | Hard | SME White Paper, table 5-1-5 |
| Gross margin to net margin regression | 0.348 × GM + 1.51, r = 0.897 | Hard | Ministry of Finance 2025 industry profit standards, 1,283 subcategories |
| Ticket for each gross margin rung | Six rungs | Inference | Derived from margin structure, not measured |
| Liquidation recovery in distribution | 20% | Industry estimate | No public data |
| Survival rate of founders with a ticket | 65% | Estimate | No measured B2B vs B2C survival data in Taiwan |
| Value uplift from a larger team | E × 1.5, survival +5 pp | Assumption | No data |
| Re-employment discount | 0.7 (age 40) / 0.5 (age 50) | Assumption | Academic findings mixed |
The first seven items are measured; the last five are estimates or assumptions.
The conclusion in section 3 that each added person drops the expected value a rung depends on the last two assumptions. Its direction is solid (opportunity cost rises linearly, business value does not), but whether the turning point comes at two or three people depends on how the 1.5 multiple is set. Partnerships at 2.52% is the one direct piece of evidence for this inference.