Layer 1 · Overview

The path forward - deals, order and a realistic payoff

A plan built as exchanges that make production pay better than taking. Design rules, the nine deals, the order to do them in, and what growth can realistically be expected.

Updated 21 Nov 2026 · 6 min read

On this page
  1. Design rules
  2. Step zero: the straight story
  3. The nine deals
  4. The order
  5. A realistic payoff
  6. What not to do
  7. Where this is weakest

If every group acts sensibly on its payoffs, the way forward is to change the payoffs. This page sets out the plan in brief: design rules, nine deals, the order of work and a realistic view of how much it can achieve.

Design rules

Before the seven design rules there is a rule zero: the ground rules for working together, with the rival test applied to every deal (where each rule changes a decision is in the strategy table).

The plan follows seven design rules, set out in full on strategy and deals. In one line: change the interest rather than ask anyone to act against it, pay in the cheapest currency (status, security, legal income, a safe exit), make the legal path cheaper than the illegal one, compensate losers once and in the open, borrow outside deadlines, build where it can be measured, and sunset every concession.

Step zero: the straight story

Before any deal, tell the public the truth: declare the economy’s difficulty, publish the losses in the banks, and replace growth targets that the forecasts do not support. Every group is deciding whether the government will still mean what it says in two years. Candour is the cheapest credibility there is.

The nine deals

# Deal The group gets The country gets
1 Pay-to-exit for bad debt A one-time settlement window; a bankruptcy route Banks that can lend again; an end to rolling rescheduling
2 Insurance pact Safety after office; caretaker rules; oversight chairs Lower stakes, longer horizons, less election violence
3 Pay for accountability The rest of the pay rise; fixed tenures Asset declarations; service standards; postings by rule
4 Cadre-to-career and legal party money Apprenticeships; open hall lists; audited party funding Less extortion; lower business costs
5 2029 export compact Tariff cuts on inputs; ports; gas; a wage rule New export lines; fewer shutdowns; labour standards the EU requires
6 Small-business compact A simple phone tax and credit record; protection from extortion for everyone A wider tax base; firms that can grow
7 Madrasa livelihoods compact Voluntary funded add-ons; a welfare fund for imams Employable graduates; access for women and safeguards against coercion
8 Barracks for budget A predictable budget; peacekeeping Out of policing and politics; published accounts
9 Migrant and diaspora compact Lower fees; skills; investment channels Higher remittances; returning skills

Each deal has a price. The bad-debt deal forgives part of what was taken and costs the public for years, and part of it may fall on large depositors through conversion of deposits into shares (Deal 1). The insurance pact applies the same fair-process guarantees to today’s leaders if they lose power, which some will see as a cost; it is not immunity. Party funding puts public money near people who behaved badly. The military keeps its commercial businesses under the barracks-for-budget deal (Deal 8). The madrasa compact puts public money and legitimacy behind religious institutions and defers family-law reform, a cost that falls mostly on women (Deal 7). Each deal is described with its price and main risk on strategy and deals.

Lessons from a deal that was tried

In April 2026 the government wrote into law a route for former bank owners to reclaim their banks by repaying the state’s money (7.5% up front and the rest within two years, with other conditions). Anger from the public, the opposition and the banking industry led to its repeal by September.1 The lessons are built into the new design: control must not go back to those who caused the losses, terms must not be payable with more bank money, and the process must be public. Repaying public support to reclaim a bank is not the same as recovering what was lost through loans that were not repaid.

The order

Four phases
  1. Now to mid-2027Stop the slide
    • Straight story: losses published, realistic targets
    • Banking commission, bankruptcy bill, IMF programme
    • Energy arrears netted, contracts published
    • January 2027 amendment: the insurance pact
  2. Mid-2027 to 2028Build the engine
    • One settlement window; weak banks resolved
    • Tariff glide path; ports; EU and Japan deals
    • Extortion and cadre pilots
    • Phone-tax and madrasa pilots
  3. 2029 to 2031Cross the graduation deadline
    • LDC graduation in November 2029; trade access secured before each market's transition ends
    • Scale what pilots proved
    • 2031 election under neutral rules
  4. 2031 to 2036Lock in
    • Peaceful handover or renewed mandate
    • Courts, police, local government deepen
    • Learning and women's work show in output
The detailed sequence, dates and scorecard are on the roadmap page.

Learning, women’s work, migration skills and cheaper cooling run alongside every phase, because they pay slowly. The full sequence is on the roadmap. The 2029 graduation date starts a clock on preferential market access: EU access continues for about three more years, other markets differ, and rules of origin and GSP+ eligibility must be secured before the transition ends.4

A realistic payoff

How far growth can move (our judgement)

Our best realistic case is slow payoff: little improvement before 2029, a recovery in the early 2030s and rough ranges set out on the roadmap. They are not model output. The government’s real-growth goal of 8.5% by FY31 is outside them; its $1 trillion goal for 2034 is a dollar-size target that also depends on inflation and the exchange rate (see the roadmap).2

This is not pessimism. A decade of 5-6% growth that is real and that people believe is worth more than a decade of 7% on paper. And the country has strong forces pulling the right way once they are no longer blocked: a young workforce, a proven export sector, remittances of $35.6 billion in FY263 and millions of small firms.

What not to do

Do not stimulate before bank balance sheets are fixed. Do not keep rescheduling bad loans. Do not expel cadres without replacing what they earn. Do not start state-directed industrial support before bank supervision works. Do not force the military out of business. Do not police private life arbitrarily. Do not open a cultural front in the next five years. Do not announce targets nobody believes.

Where this is weakest

The plan asks the governing party to help change a system it benefits from. Confidence in that is low, and the next page says why: risks and open questions. What to watch to see whether the plan is working is on what to watch.


  1. Counterpoint, The Bank Bailout Trap; Daily Star, bill to abolish the provision; BSS, repeal. ↩

  2. Government targets: Financial Express, development strategy targeting a $1 trillion economy by 2034; IMF path: Jyoti Rahman, A State of Economic Emergency. ↩

  3. Prothom Alo on the World Bank, FY26 outlook. ↩

  4. Prothom Alo, LDC graduation deferred to November 2029; EU Regulation 2026/1395, Article 17(2). ↩

Go deeper

Where this comes from

Synthesised from 5 research reports, checked against the sources cited above and reviewed before publication. How we work · All sources.