G Fund vs C Fund: when each one actually wins in retirement

1. What the G Fund actually is — and the one risk it doesn’t protect against

The TSP G Fund is unique in the U.S. retirement system. By federal statute, it invests in non-marketable U.S. Treasury securities issued specifically to the TSP, with the interest rate reset monthly to the weighted average yield of all Treasury securities with four or more years to maturity. As of May 2026, that rate sits at 4.50%. The G Fund’s principal value never drops. The chart never goes red. The worst drawdown in its 39-year history is zero.

This is what makes it psychologically irresistible — and structurally dangerous for a 30-year retirement.

The G Fund protects you against three risks that matter:

• Market risk — the C Fund’s worst single drawdown since 1988 was -55.2%. The G Fund’s was 0%.
• Interest rate risk — when rates rise, traditional bond funds fall. In 2022, the F Fund lost -18.0%. The G Fund kept appreciating.
• Credit risk — the U.S. Treasury backs every dollar.

It does not protect you against inflation risk. And for a federal retiree with a 30-year horizon, that is the risk that matters most.

2. The real-return number that changes the conversation

Nominal returns are advertising. Real returns — what your money buys after inflation — are reality.

Since its 1987 inception, the G Fund’s compound annualized nominal return has been 4.7%. Over that same period, the Consumer Price Index for All Urban Consumers averaged roughly 2.8% per year. The G Fund’s real return — what you actually gained in purchasing power — works out to 1.9% per year.

That is the central number every federal retiree needs to internalize. The G Fund grows your purchasing power at less than two percent per year, and that’s the long-run average. In the inflation spike of 2021-2023, the G Fund yielded around 2.0-3.5% while CPI hit 9.1%. During those years, the G Fund lost purchasing power every single month.

The C Fund tells a different story:

The C Fund’s gross return is roughly 6.8 percentage points higher than the G Fund’s per year. Over a 30-year retirement, that gap is the difference between leaving an inheritance and outliving your savings.

3. Why "safe" gets retirees in trouble: the sequence-of-returns problem

There’s a counterargument here that sounds reasonable: yes, the C Fund returns more on average, but it’s volatile. Volatility doesn’t matter if you’re 35 with 30 years to retirement. It matters enormously if you’re 65 and drawing down.

This is the sequence-of-returns problem, and the federal retirement community gets it backwards more often than not.

Here’s the math. If you have $1 million and never withdraw, a portfolio that loses 50% in year one and gains 100% in year two ends the same place as a portfolio that gains 100% then loses 50%. Both end at $1 million. Sequence is irrelevant.

The moment you start withdrawing, sequence becomes everything. Consider a retiree pulling $50,000 per year from a $1 million account:

• Good sequence: Year one returns +100%, taking the balance to $1.95M after the withdrawal. Year two returns -50%, leaving roughly $925K. Continues.
• Bad sequence: Year one returns -50%, taking the balance to $450K after the withdrawal. Year two returns +100%, leaving $850K. Same arithmetic, dramatically worse outcome.

Wade Pfau’s research at the American College quantifies it: approximately 77% of the final outcome of a 30-year retirement is explained by the returns of just the first 10 years. Michael Kitces’ parallel work found the correlation between first-year returns and 30-year safe withdrawal rates is only 0.21 — weak. But the correlation between the first decade’s cumulative real returns and the sustainable withdrawal rate is 0.79 — extremely strong.

Translation: get the first 10 years right, the next 20 mostly take care of themselves.

This is where most federal retirees draw the wrong conclusion. The instinct is: "If the first 10 years are critical, I should be in G Fund through that decade to protect against a crash." That instinct is wrong, and it’s wrong for a specific reason — the G Fund doesn’t lose nominal dollars, but it can still lose you the retirement. If inflation runs at 3% and the G Fund pays 4%, you’re netting 1% real on a portfolio you’re drawing 4% from. The portfolio depletes from inflation-adjusted withdrawals long before any market crash could threaten it.

The G Fund doesn’t lose nominal dollars. It can still lose you the retirement. At its recent 10-year return of 2.8%, a 4% withdrawal rate depletes $500,000 in exactly 30 years — running out at the same moment many federal retirees most need the money.

4. The math on $500,000 across 30 years

Concrete numbers. A 62-year-old federal retiree with $500,000 in TSP, drawing 4% in year one ($20,000), with the withdrawal rising 2.5% annually to keep pace with inflation. Run three allocations over a 30-year retirement using realistic recent-era returns: G Fund at 2.8% (its actual 10-year annualized return through December 2025), C Fund at 10.0% nominal (conservative long-run estimate, below its 11.5% inception average), and inflation at 2.5%.

A note on the G Fund assumption. The fund’s 39-year inception return is 4.7%, but most of that was earned in the high-interest-rate 1980s and 1990s. The recent decade (2016-2025) returned only 2.8% annualized as rates stayed historically low. Using the inception number is generous to the G Fund case. Using the recent decade is honest about what a federal retiree can expect from G Fund going forward.

Three things to notice.

The all-G-Fund portfolio depletes in year 30 — runs out exactly when many retirees still need it. At the long-run G Fund average of 4.7%, this depletion would push out to roughly year 38. At elevated inflation (3.5-4% — closer to what retirees actually experience given heavy weighting toward healthcare and housing), the G Fund would deplete at year 29-31. Either way, the buffer is thin, and the portfolio leaves no inheritance.

The 60/40 C-Fund-to-G-Fund mix grows steadily to roughly $1.44M after 30 years. This is what most retirement researchers actually recommend for a federal retiree in the early years.

The all-C-Fund portfolio nearly multiplies tenfold to $4.6M. This is the theoretical upside of accepting market volatility for the long-term return premium.

The caveat is loud: these are smooth average outcomes. Real markets don’t deliver smooth 10% annual returns. The all-C-Fund line in the real world would zigzag through drawdowns of 20-50% along the way. Most retirees can’t tolerate that volatility psychologically even when the math says it works. That’s exactly the problem the next two sections solve.

5. The retirement red zone and the rising glide path

Researchers Wade Pfau and Michael Kitces popularized the concept of the retirement red zone — roughly the 10-year window from age 60 to 70 when sequence-of-returns risk hits hardest. Three forces converge: the portfolio is at its largest absolute size, withdrawals have just begun, and there’s the least time left to recover from a bad sequence.

A market downturn during the red zone does proportionally more damage than the same downturn at any other point in the life cycle. A 30% drop on a $700K portfolio with $28K annual withdrawals starting is worse than a 30% drop on the same portfolio at age 75 with five additional years of growth and fewer remaining drawdown years.

The traditional response was a declining equity glide path — own more stocks when young, gradually shift to bonds as you approach and pass retirement. This is what target-date funds do, and what the TSP’s L Income Fund does (currently 70.7% G/F, 29.3% C/S/I as of May 2026).

Pfau and Kitces’ 2014 research challenged this. They modeled a rising equity glide path — start retirement with a lower equity allocation (e.g., 30-40% stocks), then gradually increase equity exposure through retirement. The logic: if the first decade returns are weak, you’re already in mostly bonds and protected from sequence risk; if they’re strong, you’re shifting into stocks just in time to ride the second-decade growth.

For TSP participants, the practical implication is that the L Income Fund’s static 70/30 bond/stock split may actually be sub-optimal for someone with a long life expectancy, especially a federal employee whose pension and Social Security already cover a significant portion of essential expenses.

6. The allocation that actually works in years 1-5 of retirement

There’s no single right answer here — the right allocation depends on your pension coverage of fixed expenses, your Social Security claim age, your health, and your tolerance for volatility. But there is a framework that survives most scenarios.

Step 1: Calculate your essential-expense coverage from non-TSP sources.

For a typical FERS retiree with 30 years at age 62, the FERS basic annuity covers roughly 33% of high-3 salary. Social Security claimed at 67 adds another 25-35% of pre-retirement income depending on earnings history. Together, that’s 55-65% of pre-retirement income from sources that don’t depend on TSP performance at all.

If that 55-65% covers your essential expenses (housing, food, healthcare, utilities), your TSP exists primarily for discretionary spending and inheritance. That allows for a more aggressive TSP allocation, because a 50% drawdown wouldn’t put you on cat food.

Step 2: Build a 3-5 year cash bucket in the G Fund.

For the first 3-5 years of retirement, withdrawals should come from G Fund. The amount: 3-5× your annual withdrawal need. For a retiree pulling $20K/year, that’s $60K-$100K in G Fund. This is your sequence-risk insurance — if the C Fund crashes in year two, you don’t sell at the bottom. You pull from G Fund for 3-5 years until equity markets recover.

Step 3: Keep the rest in equity-tilted allocation.

The remaining 80-85% of the portfolio sits in a C/S/I blend — typically C-heavy because most participants prefer simplicity. A common starting allocation: 60% C, 15% S, 15% I, 10% F/G blend beyond the cash bucket.

Note that even the "Conservative" row carries 60% equity exposure. Going below 40% equities in early retirement is what creates the long-term depletion risk shown in section 4.

7. The rebalancing rules that protect both sides

Allocation without rebalancing is just an opening position. The rules that make it work:

• Rebalance annually, not after every market move. Most TSP participants rebalance too often and chase volatility. Pick a date (your birthday works), look at your allocation, return it to your target.
• Refill the cash bucket from gains, not losses. In a year when the C Fund is up, sell some C Fund gains to top up the G Fund cash bucket back to 3-5 years of withdrawals. In a year when C is down, leave the cash bucket alone and let it deplete by your withdrawal — that’s its job.
• Don’t move money to G Fund after a crash. This is the single most expensive mistake federal retirees make. Selling C Fund after a 30% drop locks in the loss. The G Fund is sequence-risk insurance you build before the crash, not panic protection after one.
• TSP has a 2-transaction monthly limit. Once you’ve made two interfund transfers in a month, any subsequent transfer can only move money into the G Fund (you can’t move out). This rule alone prevents many panic-driven mistakes but doesn’t prevent the worst one — the initial post-crash flight to G Fund.

8. Where federal employees still get this wrong

Three patterns recur in FRTIB participant data and across federal retirement forums:

The default-and-forget participant. Auto-enrolled in TSP at the matching rate, defaulted into G Fund (which was the auto-enrollment fund until 2015), and never adjusted the allocation. By age 50, this participant typically has 80-100% in G Fund and a balance that’s roughly half of what a stock-tilted allocation would have produced. The fix is allocation, not contribution rate. Federal employees still in their working years can find the accumulation-phase TSP strategy at Federal Warrior’s benefits coverage ↗.

The 100% C Fund accumulator who panics at 60. Aggressive during accumulation, but as retirement approaches, the volatility starts feeling personal. Common move: shift 100% to G Fund at age 60 or 62, intending to "preserve" the gains. This eliminates sequence-risk protection (good) but also eliminates long-run growth (bad), and locks in a portfolio that will deplete from inflation-adjusted withdrawals.

The chaser. Sees the I Fund up 32.5% in 2025, moves heavily into I Fund in early 2026. Sees C Fund up 10.49% in April 2026, shifts back. The TSP’s monthly transfer limit was designed partly to slow this pattern down. The participant cost of return-chasing is well-documented: Morningstar’s annual "Mind the Gap" studies consistently show retail investor returns lag fund returns by 1-2 percentage points annually due to mistimed buys and sells.

The structural advantage of federal retirement is the pension. It exists precisely so the TSP doesn’t have to be all things — it doesn’t have to be the inflation hedge, the income generator, AND the safety net. The FERS pension is the safety net. Social Security adds inflation-indexed lifetime income. TSP can therefore afford to be the growth engine, even in retirement.

If your pension plus Social Security covers your essential expenses, the right G Fund allocation in retirement is likely 20-35%, not 60-100%. The G Fund’s job is to be a tactical reserve, not the strategic core.