Somewhere in Dhaka right now, a software company with paying customers in 40 countries is being incorporated in Wyoming. The founder is Bangladeshi, the engineers are Bangladeshi, and the salaries will be paid in taka. The company will be American, and making it so costs about $500.
The reason is Stripe. It does not support Bangladesh, and neither does PayPal in any form that lets a business here receive and withdraw revenue. Sell software to the world, and you need a payments provider, and every serious one requires an entity incorporated in a country on its list.
A private company's compliance list has become the most influential piece of corporate policy affecting Bangladeshi start-ups. That should trouble us more than the funding chart does.
Domicile is a bundle of contracts
We treat incorporation as a tax question. It is a question of what a company may sign.
A Bangladeshi private limited company cannot hold a Stripe account. It cannot issue a SAFE or a convertible note, neither of which has statutory recognition under the Companies Act 1994.
It cannot give a fund a liquidation preference backed by a forum that resolves in months. It cannot be acquired by a foreign buyer without prior Bangladesh Bank approval on transfer of control.
Each of those is a contract, and a Delaware or Singapore entity signs all of them on a Tuesday. Founders know the offshore parent will be required eventually and that doing it later is expensive, so they do it first.
The tax explanation has run out.
The Finance Act 2026 rebuilt the startup regime. A registered startup under Tk 100 crore in turnover gets a nine-year growth period, a turnover tax cut to zero across it, a loss carry-forward that survives changes in shareholding, and VAT exemption until June 2035.
For a loss-making early-stage company, the tax position in Dhaka is now at least as good as Singapore's. FEID Circular 1 of March 2026 then lets banks clear repatriation of share proceeds up to Tk 100 crore without central bank sign-off, 10 times the old ceiling, in five days.
These are serious reforms, and the people behind them deserve credit. But founders kept incorporating abroad through both.
One condition runs through the package. A startup, for these purposes, is a company incorporated under the Companies Act 1994. The relief reaches only the founders who stayed and answers a question they were not asking.
Two populations, one chart
Two groups leave for unrelated reasons. Venture-track companies flip at the first institutional round, because a fund requires an offshore parent before it wires.
Bootstrapped software companies flip in week one, because a payments provider requires a listed-country entity before it processes.
The second group is larger by headcount and invisible by construction.
When we read that start-ups raised six million dollars in the first half of 2026 and conclude the ecosystem is finished, we are reading a register of who chose Bangladeshi paperwork.
Two statutes
Foreign exchange is governed by an Act from 1947, written before the country existed. Companies are governed by an Act from 1994, written before the commercial internet.
Circulars have patched around both for two years, and a circular cannot create an instrument the underlying Act does not contemplate.
The July 2025 Share Swap circular lets shareholders swap operating company shares for offshore holdco shares with no capital leaving Bangladesh: Useful, and a bridge built toward the offshore parent rather than a removal of the reason to want one.
More improvements needed
A startup schedule inside the Companies Act would do most of the work. Statutory recognition for convertible instruments. Primacy for shareholder agreements over articles, so preferences and drag-along bind as written.
A commercial forum with a fixed clock, because investors are pricing years and months, would change that price. And publication of Bangladeshi-founded funding beside Bangladesh-domiciled funding, because a country cannot tax what its statistics decline to see.
I have raised money into a Bangladeshi cap table and into a Singapore one. The difference was never conviction about where to build. It was that one legal vocabulary had a word for the instrument the investor held.
Six million dollars is a number about paperwork. The companies are here.
Hussain M Elius co-founded Pathao and founded Wind. He is the author of "Moving Bangladesh: The Origin Story of Pathao."
