Intelligence
Research, insights, data and media on African institutional capital.
Intelligence overviewAnalysis
Solutions
Three practices, one obligation — and the platform that connects African asset owners to capital, opportunities and peers.
Solutions overviewOur practices

Decision Alpha · Inaugural Cohort
One framework that every module, case, visit and tool feeds. It works at three levels and moves from diagnosis to discipline.
Decision Alpha
Biases are not a failing of intelligence. They are the default settings of every human mind, and they get stronger under time pressure, public scrutiny and large sums. The question is not whether a board is exposed to bias, but whether its process catches bias before it becomes a decision.
The behavioural risks are common to every asset owner. What differs is who the board answers to, what constrains it, and where pressure enters. The programme works through each level at which bias enters a decision: the individual member, the committee and the institution.
The Decision Alpha framework
| Level | The question | Where biases enter | What the programme builds |
|---|---|---|---|
| Individual | How does each member judge? | Overconfidence, anchoring, recency, loss aversion | Calibration, probabilistic thinking, base rates |
| Committee | How does the group decide? | Groupthink, authority, herding, deployment pressure | Independent views before debate, red-teaming, chairing for challenge |
| Institution | How does the institution monitor and learn? | Escalation of commitment, planning fallacy, hindsight | Stage-gates, kill criteria, decision logs, post-decision review |
Six behavioural risks
| Bias | How it shows up | Where exposure sits | Safeguard |
|---|---|---|---|
| Recency and anchoring | Last year’s return becomes the benchmark for members, ministers or management targets | Declared rates, bonuses, budget transfers; strategy targets | Rolling five- and ten-year returns against the real-return objective; realised separated from fair-value gains |
| Overconfidence | Narrow forecasts in asset classes the institution has little history in | Infrastructure, private equity, fund-of-funds | Reference-class forecasts; ranges, not point estimates |
| Escalation of commitment | Funding a delayed project because of what has already been spent | Real estate, legacy land, development projects | Stage-gates with kill criteria set before approval; independent review at each gate |
| Planning fallacy | Budgets and dates that assume nothing goes wrong | Real estate and infrastructure pipelines | Contingency sized from comparable projects; delay costs priced into the return case |
| Herding and authority | Following government, peers or the most senior voice over the evidence | National project requests; regional deals | Written independent challenge; members state views before the most senior voice |
| Deployment pressure | Approving deals to put cash to work and hit asset targets | Fast-growing institutions in shallow markets | No deployment targets in IC papers; a hurdle rate and a do-nothing option in every paper |
Seven habits for the board’s charter
| Safeguard | Counters | How it enters the board’s process |
|---|---|---|
| 1. Pre-mortem in every major paper | Overconfidence, planning fallacy | A mandatory section in the paper template above an agreed threshold |
| 2. Kill criteria and stage-gates set at approval | Escalation of commitment, sunk cost | A missed gate returns the project to committee automatically |
| 3. Realised versus fair-value returns; rolling five- and ten-year figures | Recency, anchoring, outcome bias | Standard format for all performance reporting |
| 4. Reference-class forecasts and ranges | Overconfidence, planning fallacy | Required for cost, timeline and return assumptions |
| 5. Independent challenge on infrastructure, private equity and government-linked deals | Herding, authority, deployment pressure | An internal red team or external reviewer reports to committee |
| 6. Decision log with rationale, assumptions and dissent | Hindsight and outcome bias | Reviewed annually against outcomes |
| 7. Views recorded before discussion | Authority bias, groupthink | Written positions on major approvals before debate opens |
Programme shape
Two days in residence, three days inside London institutions, and a follow-up that measures whether behaviour changed.
Delegations of three to five from six to eight African asset owners, up to 30 participants, mixing types of asset owner.
A board self-assessment, a calibration quiz, a knowledge check and a briefing paper. Each delegation submits one live or recent decision as its case, agreed at a scoping call.
How individuals decide, then how markets and committees behave, with specialist faculty bringing their own cases into the room.
Three a day, one of them an asset owner. At every visit participants do something and leave with a tool. Each day ends with the debrief “What will we adopt?”
Each delegation drafts its seven safeguards for board adoption, and the calibration quiz is re-run.
A completion report within ten working days, a virtual follow-up and self-assessment re-run at 90 days, and, as an option, a review of two live board decisions at six months.
Request the prospectus, the indicative fees and the Investment Decision Charter outline, tailored to your institution.
Curriculum
Days 1–2 take participants through all sixteen modules in residence. Days 3–5 put them into practice at hosted visits.
| Module | The question it answers | What is covered | Tool |
|---|---|---|---|
| M1 Calibration and self-assessment | How well do we know what we don’t know? | Calibration and self-assessment results; confidence ranges; the gap between certainty and accuracy | Personal calibration score |
| M2 Behavioural science foundations | Why do sensible people decide badly? | Prospect theory, loss aversion, heuristics, bias versus noise; why annual reporting cycles make boards judge long-term investments on one-year results | Bias glossary for board use |
| M3 Biases at board level | Where does bias hide in a board pack? | Syndicates review a UK committee pack, then an anonymised African asset owner’s paper; where “to note” replaces challenge | Paper review checklist |
| M4 Making sense of a strong year | Was it skill or luck? | Realised versus fair-value returns; outcome bias; how a strong year frames the expectations of members, policyholders, ministers and management | Return reporting format |
| M5 Risk, uncertainty and probabilistic thinking | What happens if three things go wrong at once? | Base rates, expected value, scenarios; a liquidity shock against each syndicate’s own liabilities, combined with a currency reversal and a rate rise | Liquidity scenario template |
| M6 Challenging the business case | What do comparable projects tell us? | Reference-class forecasting on a mock infrastructure PPP; sponsor optimism | Reference-class worksheet |
| M7 Pre-mortem and red-teaming | It is five years on and this failed. Why? | Live pre-mortem; structuring a red team; turning failure reasons into gates | Pre-mortem protocol |
| M8 Bubbles, herding and momentum | What does a bubble look like from inside? | From the South Sea Bubble of 1720 to 2008 and frontier-market cycles | Herding warning signs |
| M9 Reading market signals | Which signals inform, and which distort? | Equity rallies, currency moves and rate rises; why a rate rise lifts new bond yields but cuts the value of bonds held | Signal classification grid |
| M10 Behavioural portfolio discipline | What commits us before temptation arrives? | Investment beliefs and strategic allocation as commitment devices; liability, solvency-capital and regulatory constraints; pacing deployment; hiring and firing managers without chasing performance | Draft investment beliefs |
| M11 Investment committee simulation | Would we approve this, under pressure? | A national infrastructure proposal as a full committee paper: one delegation presents, a different type of asset owner red-teams, the board decides, with views recorded before debate | Challenge log |
| M12 Independent judgement and constructive challenge | Whose interest does each director serve? | Groupthink, authority, chairing for challenge; serving beneficiaries, not the bodies that nominate or sponsor each director | Chairing guide |
| M13 AI, markets and investment decisions | Does AI reduce bias or add to it? | AI-driven trading and herding; model risk; AI as analyst and as independent challenger; governing its use | AI use principles |
| M14 Projects after approval | When should we stop? | Escalation of commitment, stage-gates and assurance; early-warning reporting | Stage-gate and kill-criteria checklist |
| M15 The Investment Decision Charter | What will we change on Monday? | Drafting the seven safeguards from the week’s Charter log | Draft Charter |
| M16 Action plan and close | Who owns what, by when? | Board knowledge and skills plan; calibration re-run; commitments, owners and the 90-day review date | Action plan |
Before London
Sent two weeks ahead. It gives each participant a personal baseline, gives faculty a picture of each board, and turns the first morning into a discussion of real results.
| Element | What it is | Why it matters |
|---|---|---|
| Board self-assessment | Thirty confidential statements on role clarity, paper quality, meeting behaviour, risk appetite and early warning of problems, on a four-point scale with no neutral option | Opens Day 1 and is re-run at 90 days, so change is measured |
| Calibration quiz | Ten questions on each delegation’s own market and the wider region, answered as 90% confidence ranges | Most people’s ranges are too narrow; seeing it in their own results is the fastest route to humility about forecasts |
| Knowledge check | A short test on investment, risk and markets, modelled on the knowledge assessments UK pension boards take | Sets the technical level of each session |
| One decision to revisit | Each member names one past decision of their institution they would now make differently | Grounds the week in the participants’ own experience |
| Institution case | Each delegation submits one live or recent decision, anonymised as it chooses | Becomes its case for the committee simulation and peer red-team |
| Briefing paper | Eight to ten pages on the behavioural risks facing asset owners’ boards, with notes on each participating institution’s sector and country | About 30 minutes’ reading; a common starting point for the cohort |
| Scoping call | A call with each delegation’s sponsor and company secretary | Agrees cases, sensitivities and what may be named |
The case library
Worked examples come from public sources; constructed cases give each type of asset owner a decision built on its own balance sheet; each delegation brings its own. Nothing is named without the institution’s approval.
| Case | Type | What it teaches | Modules |
|---|---|---|---|
| The record year: a national social security fund declares about twice its ten-year average return | Worked example | Recency, anchoring, outcome bias; realised versus fair-value returns; how quickly gains can reverse | M4, M9 |
| The liquidity shock: a pension fund opens mid-term access to members’ savings | Worked example | Policy risk to cash needs; illiquid commitments against a maturing membership | M5 |
| The national project request: an infrastructure financing request modelled on an expressway PPP | Constructed case | Sponsor optimism, deployment pressure, authority bias | M6, M7, M11 |
| The project that would not die: a land and housing project unresolved after almost two decades | Worked example | Escalation of commitment, planning fallacy, monitoring failure | M14 |
| The UK committee pack: manager presentations, a pooling paper and a private equity reinvestment report | Benchmark | What a mature committee is asked to decide, and where “to note” replaces challenge | M3, M10 |
| The investment strategy statement: beliefs, pooling and pacing of private-market commitments | Benchmark | Investment beliefs as a commitment device | M10, M15 |
| The collateral call: UK pension funds in the September 2022 gilt crisis | Benchmark | How liquidity risk hides inside a sound-looking strategy | M5, M9 |
| The capital-constrained allocation: an insurer considers moving surplus assets into private credit | Constructed case | Regulatory anchoring, reaching for yield, capital-adjusted returns | M5, M10 |
| The national champion: a sovereign fund is asked to co-invest in a priority domestic company | Constructed case | Authority bias, home bias, the counterfactual test | M11, M12 |
| The loss year: a reserve management committee weighs a new asset class after mark-to-market losses | Constructed case | Loss aversion, reputational asymmetry, status-quo bias | M2, M5 |
| The spending rule: an endowment board is pressed to raise spending after a strong year | Constructed case | Recency, present bias, intergenerational fairness | M4, M10 |
| Your decision: one live or recent decision from each delegation | Institution-supplied | Applying every tool of the week to a real decision, with peer challenge | M7, M11, M15 |
The research base
| Finding | What it means for a board | Source |
|---|---|---|
| Plan sponsors hire managers after strong runs; the managers hired do not then beat the managers fired | Recent performance is a weak guide to the future; hire and fire by rules | Goyal and Wahal, 2008 |
| Large projects overrun cost and time far more often than not | Forecast from comparable projects, not from the sponsor’s plan | Flyvbjerg, 2006 |
| People add money to failing courses of action they chose themselves | Separate those who approve from those who review | Staw, 1976 |
| Losses weigh more than gains, and frequent evaluation raises the pain | Short-horizon review pushes boards to the wrong level of risk | Benartzi and Thaler, 1995 |
| Imagining a failure has already happened surfaces more, and more specific, risks | Build a pre-mortem into every major paper | Klein, 2007 |
| Judges given the same case reach widely different answers | Measure the noise in the committee, not just the bias | Kahneman, Sibony and Sunstein, 2021 |
| Better-governed pension funds tend to perform better | Process quality is an investment asset, not an overhead | Clark and Urwin, 2008 |
Designed for a mixed cohort
A single-institution programme teaches a board about itself. A cohort of different asset owners adds peers who face the same biases from a different balance sheet, and who are not bound by each other’s hierarchy.
Calibration results are compared across delegations and institution types, anonymised, so each board sees where it sits against its peers.
Every delegation scores the same proposals independently. The spread shows how much of a decision depends on who happens to be in the room.
In the committee simulation, a different type of asset owner challenges your proposal: an insurer through a capital lens, a reserve manager through a liquidity lens.
Where material is type-specific (liabilities, solvency capital, fiscal rules, spending rules) delegations work with like institutions, then report back.
Each delegation maps its investment beliefs against its own regulator’s rules and its market’s depth, so safeguards fit local reality.
Chatham House Rule throughout; institution-supplied material stays in the cohort; nothing is named without approval.
Shared principles. The cohort agrees a short set of Deciding Well Principles for African Asset Owners, which Asset Owners Africa can publish as a reference for the sector.