CDC vs DC vs DB: What 80 years of data tells us
This content is AI generated, click here to find out more about Transpose™.
For terms of use click here.
In this CDC investment series, we explore how CDC schemes work in practice, from core design features to real world outcomes across different market conditions. Chapter 2 looks at 80 years of data to show how CDC compares with DC and DB.
Explore Chapter 1: Why investment is critical to CDC
This analysis uses 80 years of real market data to explore how Collective Defined Contribution (CDC) pension schemes perform across very different economic environments, and how their outcomes compare with individual DC and traditional DB arrangements.
Across all periods considered, well-designed CDC schemes prove resilient to radically different market conditions. In our modelling, CDC outperformed individual DC with annuity purchase in every period tested. This is not because CDC eliminates risk, but because of how risk is shared and managed within the scheme.
In particular, CDC benefits from three structural features:
- Sustained exposure to growth assets, allowing members to participate in long-term equity returns without being forced to crystallise losses at retirement
- Intergenerational risk sharing, which smooths the impact of sharp changes in yields and market conditions
- Adjustable indexation, enabling the scheme to absorb short-term volatility while still capturing upside when markets recover
CDC is not immune to adverse conditions. The period from 1963 to 1983 characterised by high inflation and weak real returns was especially challenging. During this time, benefit increases lagged inflation and outcomes were weaker. However, even in this extreme environment, nominal cuts were avoided and younger members, who later benefited from the recovery, were better positioned to absorb volatility.
While the historical periods analysed are unlikely to be repeated exactly, they provide a useful lens through which to understand how CDC schemes may behave under a wide range of future market scenarios.
Our approach
Our analysis uses actual historic returns and market conditions to model outcomes from different types of pension arrangements over time.
Historic interest rate and inflation data was used to calculate total asset returns, contribution rates and actual outcomes for uplifts, cuts and pension increases. Mortality was assumed to be the same across all time-periods. More details of the approach is given within Modelling approach section.
From this we derive the “replacement ratio”, our key measure of pension outcomes at retirement.
Replacement ratio is defined as ratio of pension at retirement to final salary. It ignores state pension and assumes no tax-free lump sum. A replacement ratio of 37% means that for a salary of £100 p.a., the member receives a pension of £37 p.a. after retirement.
The replacement ratio only captures pension size at retirement, it does not include any potential upside or downside after retirement.
Findings
1. CDC is resilient across very different market environments
We first compare replacement ratios across different pension designs using identical contribution patterns and historical market returns for a member who is 40 years old at the beginning of each period, retiring at 60 at the end of each period. The results below show how outcomes vary across four distinct economic regimes spanning the past 80 years.
Across all four periods modelled, CDC delivers higher replacement ratios than individual DC with annuity purchase, while also showing less sensitivity to market regime. The contrast is particularly clear in more challenging environments, such as 2003–2023, where poor timing at retirement materially reduces DC outcomes.
DB outcomes are stable by construction, reflecting the fact that risks are absorbed elsewhere. CDC sits between DC and DB: it does not eliminate risk, but it avoids crystallising losses at a single point in time.
2. Sustained exposure to growth assets is the primary driver of outcomes
A consistent feature of CDC across all periods is its ability to maintain exposure to growth assets throughout retirement. Unlike DC annuity purchase, which converts assets into fixed income at retirement, CDC allows outcomes to adjust over time as markets evolve. This is particularly evident in the 1943 – 1963 period where strong market performance was passed on to members for the entire period.
This feature underpins CDC’s relative resilience: poor market conditions at retirement do not permanently lock in lower pensions, while subsequent recoveries can still be reflected in benefit levels.
3. Outcomes within CDC vary by age and horizon
Outcomes also differ by the age of the member accruing benefits. Younger members benefit from cheaper accrual in the period of depressed market performance. This is illustrated by the chart below which shows comparison of replacement ratios for 20- and 40-year-old members.
In all time-periods the younger member accrues a significantly higher pension than the older member. This is primarily due to accrual being cheaper for younger members because they have a longer horizon.
Beyond this, scheme experience impacts accruing benefits in several conflicting ways. For example, in a good year in a CDC scheme the following things generally happen:
- Current pension increases are up. This benefits everyone as increases apply to all entitlements
- Expected future pension increases are up. This benefits members in retirement or close to retirement since their pension goes up faster than previously thought.
- However, the cost of buying (accruing) pension is also up. This is because one pound of pension now carries higher increases and thus costs more to acquire.
The opposite occurs in poor conditions- so pension increases fall, but the cost of accrual becomes cheaper.
In general, younger members are more sensitive to changes in the price of accrual, while older members care more about protecting the current purchasing power of their pension.
This can be seen in the analysis, where:
- The accruing (younger) member performed very strongly in the final period relative to other periods
- And surprisingly well in the second period (high inflation / stagflation) with the 3rd period ending up being the worst period.
This is because the second and fourth periods featured prolonged periods of “cheap” accrual due to weaker scheme performance, which allowed younger members to build up significantly larger entitlements.
4. CDC adjusts pensions over time rather than locking in outcomes
The way CDC responds to market conditions during retirement is illustrated by how pensions in payment evolve in real terms.
The graph below tracks the pension a member retiring from a CDC scheme at the start of the period would receive and shows their pension in real terms (net of inflation) over 20 years of retirement. For simplicity the real pension is expressed as an index.
In strong return environments, pensions increase materially in real terms. In more challenging periods, particularly 1963–83, increases fail to keep pace with inflation and real pensions fall. Crucially, however, CDC avoids large nominal cuts and allows pensions to recover when conditions improve.
This illustrates the core trade-off in CDC: variability in pension increases replaces the risk of permanently locking in poor outcomes at retirement.
Conclusion
Using 80 years of market history, this analysis highlights how different pension designs respond to very different economic environments. CDC does not eliminate investment risk, nor does it guarantee outcomes. Instead, it changes how risk is shared and how outcomes adjust over time.
Relative to individual DC with annuity purchase, CDC avoids locking in poor outcomes at a single point in time and allows members to continue participating in long-term growth during retirement. Compared with purely individual arrangements, it spreads the impact of adverse market conditions across members and over time, reducing the severity of poor timing for any one cohort.
These features come with trade-offs. Outcomes within CDC vary by age and market experience, and pensions in payment may fluctuate in real terms. Whether this is viewed as desirable depends on how fairness is defined: as individual ownership of outcomes, or as collective sharing of risk.
The experience of the past 80 years does not predict the future, but it does illustrate the structural strengths and limitations of different pension designs. For policymakers, trustees, and employers considering CDC, the key question is not whether risk can be removed, but how it should be shared, and whether members are comfortable with the form that sharing takes.
Subscribe to our thinking
Get relevant insights, leading perspectives and event invitations delivered right to your inbox.
Get started to select your preferences.
We divided the last 80 years into four equal periods that roughly align with major shifts in inflation, interest rates and asset returns.
| Period | Description | Real equity return | Interest rate | Inflation |
| 1943-1963 | Post-war boom, strong equities, moderate inflation | 12.0% | 4.2% | 3.1% |
| 1963-1983 | Stagflation, high inflation & rates, weak equities | 2.8% | 10.7% | 9.0% |
| 1983-2003 | Moderate inflation, falling rates, strong bonds and volatile equities | 5.4% | 8.2% | 3.1% |
| 2003-2023 | Low rates & inflation, strong equities (despite crises) | 7.1% | 2.8% | 2.8% |
We tracked three representative members across each time-period:
- Age 20 (accrues for 20 years, then defers until age 60)
- Age 40 (accrues 20 years, retires at 60)
- Age 60 (already in payment at the start)
Members are assumed to join an established CDC scheme with a stable member population. The key features of the schemes considered are:
| CDC | DC | DB | |
| Contribution rate | 10% of salary | 10% of salary | Accrual rate met regardless of cost |
| Investment strategy | Progressive member-level de-risking age 60→90 from Growth to Matching | 100% Growth → full de-risk 10 yrs pre-retirement |
Not modelled (promises are met regardless of cost) |
| Benefit promise |
None Target increase: CPI+0%. Cap: CPI+2%: Floor: 0% (no increases) |
Individual pot | Guaranteed 1/80th accrual (25% of salary after 20 yrs) |
| Adjustment | Cuts or uplifts applied outside the bounds if needed to keep scheme balanced | None |
None – cost absorbed by sponsor
|
| At retirement aged 60 with no lumpsum |
Pension paid directly from scheme |
Compulsory annuity purchase at Gilt - 0.2% |
Guaranteed DB pension |
Scheme and market performance
Post War years – 1943 - 1963
The scheme performed very strongly, passing on robust investment returns to members. Low inflation relative to strong equity performance resulted in numerous benefit uplifts throughout the period.
Inflationary 60 and 70s
This was the most challenging period. The scheme awarded below-inflation increases for most of the period — with particularly weak performance between 1975 and 1980 — driven by lacklustre returns and very high inflation.
Volatile 80s and 90s
Performance was generally positive. Despite significant market volatility, the scheme avoided cuts and awarded several benefit uplifts to reflect strong equity performance. It broadly maintained and increased members’ purchasing power.
The new millennium
The period proved relatively benign overall for the scheme. It weathered the 2008 financial crisis and returned to granting above-inflation increases later in the period. High inflation towards the end was generally offset by healthy equity performance.
