TSP Funds Explained: A 2026 Complete Guide to G, F, C, S, and I
TSP funds explained: five funds, two stocks-versus-bonds decisions, one of the cheapest retirement plans in the world. Here’s what each TSP fund holds, how it has performed through May 2026, and the mix that fits each career stage.
1. The five funds at a glance
The Thrift Savings Plan offers five individual investment funds and eleven Lifecycle (L) Funds. The five core TSP funds — G, F, C, S, I — are the building blocks. Everything else, including the L Funds, is a recipe made from those ingredients. Here is where each TSP fund stands as of May 13, 2026:
| Fund | What it holds | YTD 2026 | 1-year | 10-year | Since inception | Expense ratio |
|---|---|---|---|---|---|---|
| G | Special-issue Treasury securities | +1.5% | 4.4% | 2.8% | 4.7% | 0.048% |
| F | Bloomberg US Aggregate Bond Index | 0.0% | 5.3% | 1.7% | 5.3% | 0.061% |
| C | S&P 500 (large-cap US) | +9.2% | 28.0% | 15.7% | 11.5% | 0.049% |
| S | Dow Jones US Completion (small/mid-cap) | ~-1% | ~18% | 11.3% | 9.6% | 0.051% |
| I | ACWI IMI ex-US-ex-China-ex-HK | +14.2% | 37.5% | 10.4% | 6.5% | 0.054% |
The expense ratios alone deserve a moment. The most expensive TSP fund costs 5.4 cents per $1,000 invested per year. The average 401(k) plan in the United States runs 50 to 150 cents per $1,000 — roughly 10 to 30 times more. Over a 30-year federal career on a six-figure balance, the difference is tens of thousands of dollars in compounded savings that never leave the account as fees.
This cost advantage is the structural reason the TSP keeps producing record numbers of millionaire participants — 194,722 as of January 1, 2026, dropping to roughly 185,000 by April after the Q1 market correction. The funds themselves are unremarkable index trackers. The fees are what compound differently.
The chart shows the 1-year returns across all five funds. The I Fund’s 37.5% leads the lineup — a striking reversal of the previous decade when international stocks dramatically underperformed U.S. stocks. The G Fund’s 4.4% is the only "guaranteed-positive" line on the chart, but it’s also the worst real-return performer when inflation is factored in. The point isn’t that one fund is "best" — it’s that each fund does a different job, and ranking them only makes sense once you know what job you’re trying to fill.
Notably absent from the TSP lineup: emerging markets beyond what’s in the I Fund, REITs, commodities, high-yield bonds, individual stocks, and ETFs. The Mutual Fund Window covers some of these — at significant cost — but the core TSP is intentionally narrow. That narrowness is a feature, not a bug, for participants who’d otherwise overtrade.
2. G Fund — the government securities anchor
What it holds: Non-marketable U.S. Treasury securities issued specifically to the TSP. These aren’t traded on any open market. They exist only inside the plan.
How the interest rate is set: By federal statute, the Treasury Department resets the G Fund rate monthly to the weighted average yield of all U.S. Treasury securities with four or more years to maturity. The current rate, set on the last business day of the prior month, is 4.50% as of May 2026.
The unique feature: Long-term yields, short-term liquidity. Normally, you’d have to lock up your money in long-dated Treasuries to earn long-term rates. The G Fund pays the long-term yield but lets you withdraw on any business day with no principal risk and no interest-rate risk. There’s no equivalent product available to private-sector workers.
Long-run performance: Since its April 1, 1987 inception, the G Fund has earned an annualized 4.7% nominal. That sounds modest, but with zero volatility and a 0% worst drawdown, it has actually outperformed many "low-risk" alternatives like money market funds, short-term bond funds, and most CDs across most periods.
The honest weakness: Inflation. The TSP itself acknowledges this on tsp.gov: "The G Fund is subject to the possibility that your investment will not grow enough to offset the reduction in purchasing power that results from inflation." Subtracting average CPI (about 2.8% over the G Fund’s life) from the 4.7% nominal return yields a real return of roughly 1.9% per year. During the 2021-2023 inflation spike, the G Fund yielded 2-3.5% while CPI peaked at 9.1% — every G Fund participant lost purchasing power for nearly two years.
For more on when the G Fund’s apparent safety becomes a long-term liability, see G Fund vs C Fund: when each one actually wins in retirement.
3. F Fund — the bond fund nobody talks about
What it holds: A diversified U.S. bond portfolio tracking the Bloomberg U.S. Aggregate Bond Index — roughly 13,000 investment-grade bonds spanning U.S. Treasuries, corporate bonds, mortgage-backed securities, and asset-backed securities.
Why it’s the most overlooked fund: The F Fund is mathematically duplicative for most participants who already hold G Fund. The G Fund offers similar (or better) returns to short-and-intermediate Treasuries without principal risk. The F Fund adds credit risk and interest-rate risk in exchange for a small additional yield from corporates and mortgages.
Recent performance:
| Period | F Fund | G Fund | Difference |
|---|---|---|---|
| YTD 2026 | 0.0% | +1.5% | G wins |
| 1-year | 5.3% | 4.4% | F wins slightly |
| 5-year | 0.3% | 3.7% | G dramatically wins |
| 10-year | 1.7% | 2.8% | G wins |
| Since inception (1988) | 5.3% | 4.7% | F slightly wins |
The five-year picture is brutal for the F Fund. From 2021 to early 2023, the F Fund lost roughly 18% as interest rates rose and bond prices fell. The G Fund, by contrast, kept appreciating. This isn’t a failure of the F Fund — it’s how all aggregate bond funds work when rates rise. But it makes the case that for a TSP investor, the G Fund is usually the better "safe" choice unless you specifically want exposure to credit risk and the corporate/mortgage segments.
The F Fund is useful in two specific situations: when you’ve maxed out other diversification options inside TSP and want bond exposure beyond Treasuries, or when you’re trying to match a target-date fund’s allocation to roll over an outside 401(k). For most career federal employees, the G Fund handles the safe-assets role more efficiently.
4. C Fund — the S&P 500 workhorse
What it holds: Shares of the 500 largest U.S. companies — Apple, Microsoft, Nvidia, Alphabet, Amazon, Meta, Berkshire Hathaway, and so on — weighted by market capitalization. It tracks the S&P 500 Total Return Index.
Why most federal employees end up here: The C Fund is the workhorse of the TSP. It’s the most-allocated-to individual fund in the plan, and it accounts for the bulk of the wealth held by TSP millionaires. As of May 13, 2026, the C Fund has:
- YTD 2026 return: +9.2% (after a rough Q1 that ended at -4.34% YTD, recovered by April’s +10.49% single-month return)
- 1-year return: 28.0%
- 5-year annualized: 14.2%
- 10-year annualized: 15.7%
- Since inception (1988): 11.5%
A $1,000 investment at the C Fund’s January 1988 inception would be worth approximately $64,674 today — a 64-fold increase over 38 years.
The real cost: volatility. The C Fund has an annualized standard deviation of 17.8%, meaning roughly two-thirds of annual returns fall within a range of plus-or-minus 18 percentage points from the average. The fund’s worst peak-to-trough drawdown since inception was -55.2% during the 2008-2009 financial crisis. Anyone telling you "the C Fund always goes up" is missing the 18-month period when it dropped to less than half its starting value.
Why the volatility usually doesn’t matter for accumulators: If you’re 30 years from retirement, a 55% drawdown is mathematically temporary because you’re buying more shares every payday during the decline. Federal employees who kept contributing through 2008-2009 ended up with substantially more wealth by 2020 than those who panicked and moved to G Fund near the bottom.
The C Fund’s worst drawdown since 1988 was 55.2%. The C Fund’s 38-year annualized return is 11.5%. Both numbers are true at the same time. The discipline is staying invested when the first number is showing up on your statement.
5. S Fund — the small and mid-cap completion
What it holds: Shares of every U.S. company not in the S&P 500 — roughly 2,500 small and mid-cap companies tracking the Dow Jones U.S. Completion Total Stock Market Index. Think of it as "everything in the U.S. stock market that isn’t already in the C Fund."
Why it exists: Small and mid-cap U.S. stocks have historically delivered slightly higher long-term returns than large caps, with proportionally higher volatility. Owning both C and S Funds together gives you exposure to the entire U.S. stock market.
Recent performance:
- YTD 2026: roughly -1% through May (the S Fund had a rough Q1 at -1.22% YTD then a strong April at +9.96%, finishing close to flat through mid-May)
- 1-year return: approximately 18%
- 10-year annualized: 11.3%
- Since inception (May 2001): 9.6%
- Annualized standard deviation: 22.0% (the highest of any TSP fund)
A $1,000 investment at the S Fund’s May 2001 inception would be worth approximately $9,971 today.
The trade-off: Higher volatility than C, but no guaranteed return premium. Over the last 10 years, the S Fund’s annualized 11.3% is actually below the C Fund’s 15.7% — large-cap U.S. stocks have outperformed small-and-mid caps for most of the post-2015 period. Whether the S Fund’s long-run premium reasserts itself remains a perennial debate in retirement planning.
A common diversification rule of thumb: weight C and S funds in proportion to their relative market capitalization. As of 2026, the S&P 500 represents roughly 75-80% of total U.S. stock market value, and the S Fund’s index covers the remaining 20-25%. So an "all U.S. stocks" TSP allocation looks roughly like 78% C Fund and 22% S Fund. Many participants simplify to 80/20.
6. I Fund — international, after the 2024 overhaul
What it holds (as of 2026): Shares of approximately 5,500 companies across 44 countries — 21 developed markets plus 23 emerging markets. Notable exclusions: China and Hong Kong (excluded by congressional pressure, including the S.1665 Prohibiting TSP Investment in China Act).
The fund tracks the MSCI ACWI IMI ex-USA ex-China ex-Hong Kong Index.
Why this matters — the 2024 overhaul: The I Fund used to be much narrower. Prior to 2023, it tracked the MSCI EAFE Index, covering only about 790 stocks in 21 developed-market countries (mostly Japan and Europe), with no Canada and no emerging markets at all. In 2024, the FRTIB switched to the current broader index, dramatically expanding the I Fund’s diversification.
The result: federal employees now get genuine global stock exposure inside the TSP, instead of a Europe-and-Japan-only fund.
2026 performance:
- YTD 2026: +14.2% (the strongest fund in the lineup year-to-date)
- 1-year return: 37.5%
- 10-year annualized: 10.4%
- Since inception (May 2001): 6.5%
The I Fund’s strong 2025 and early 2026 performance reversed a decade-long pattern of international underperformance vs. U.S. stocks. From roughly 2010 to 2023, U.S. stocks dramatically outperformed international, leading many TSP participants to underweight or skip the I Fund entirely. Whether the current outperformance continues is the open question of 2026 portfolio allocation.
The currency factor: The I Fund is not currency-hedged, meaning returns are influenced by U.S. dollar movements. When the dollar weakens (as it has in parts of 2025-2026), international holdings translate to more dollars and the I Fund gets a boost. When the dollar strengthens, returns drag.
A truly globally diversified international fund (like Vanguard’s VXUS) holds roughly 7% in China and 1.4% in Hong Kong. The TSP I Fund excludes both for political reasons. Federal employees who want full international exposure may need to allocate a small slice outside the TSP — typically through an IRA — or accept the exclusion. The gap is meaningful but not catastrophic.
7. L Funds — the auto-allocation option
The Lifecycle Funds aren’t separate investments — they’re pre-built recipes that hold the five individual funds in specific proportions, automatically shifting toward more bonds as the target year approaches.
Eleven L Funds as of 2026:
- L Income — for participants already withdrawing or near withdrawal
- L 2030 — target withdrawal 2028-2032
- L 2035 — target 2033-2037
- L 2040 — target 2038-2042
- L 2045 — target 2043-2047
- L 2050 — target 2048-2052
- L 2055 — target 2053-2057
- L 2060 — target 2058-2062
- L 2065 — target 2063-2067
- L 2070 — target 2068-2072 (inception July 2024)
- L 2075 — target 2073+ (inception June 2025)
What they cost: Same underlying expense ratios as the individual funds, weighted by allocation. No extra "fund of funds" fee. This makes them genuinely free auto-rebalancing.
What’s in them:
| L Fund | G/F % | C % | S % | I % | Equity total |
|---|---|---|---|---|---|
| L Income | 70.7% | 11.8% | 3.5% | 13.9% | 29.2% |
| L 2030 | 51.5% | 22.3% | 6.4% | 19.7% | 48.4% |
| L 2040 | 26.6% | 33.6% | 9.9% | 29.9% | 73.4% |
| L 2050 | 9.9% | 41.6% | 12.7% | 35.8% | 90.1% |
| L 2060+ | 0.7% | 44.4% | 13.9% | 41.0% | 99.3% |
Who should use them: Participants who don’t want to manage allocation themselves, especially auto-enrollees under 30 who don’t have a strong view on stocks vs. bonds. The L Funds are calibrated for the average federal employee — someone who may not have a pension at all in retirement.
The hidden assumption: L Funds assume your TSP is the primary leg of retirement. For a FERS employee with a 30-year pension, the L Income Fund’s 70% bond allocation may actually be too conservative, because the FERS pension already covers a substantial fraction of essential expenses. Many retirement researchers argue that federal retirees with strong pension coverage can afford an equity-tilted allocation more aggressive than what the L Income Fund provides.
For more on TSP allocation in retirement specifically, see G Fund vs C Fund: when each one actually wins in retirement.
8. The Mutual Fund Window: when it’s worth $132 a year
In June 2022, the TSP added the Mutual Fund Window (MFW) — an option to invest up to 25% of your TSP balance in roughly 5,000 outside mutual funds from Vanguard, Fidelity, T. Rowe Price, and about 300 other fund families.
Almost nobody uses it. As of February 2026, only about 9,000 of 7 million TSP participants had funded a Mutual Fund Window account. That’s roughly 0.13% adoption.
Why so few: Fees and complexity.
The fees stack:
- $95 annual maintenance fee to the brokerage provider
- $55 annual administrative fee to the FRTIB (to ensure non-MFW users don’t subsidize the option)
- $28.75 per trade (buy and sell)
- Plus the expense ratio of each mutual fund you actually invest in
The fee math:
| Total MFW balance | Annual fixed fees | Fee drag % |
|---|---|---|
| $10,000 | $207.50 | 2.08% |
| $25,000 | $207.50 | 0.83% |
| $50,000 | $207.50 | 0.42% |
| $100,000 | $207.50 | 0.21% |
| $250,000 | $207.50 | 0.08% |
At the $10,000 minimum allocation (the entry point), MFW costs you roughly 2% annually before the underlying mutual fund’s expense ratio. That’s higher than 99% of all mutual funds in the U.S. retail market. The fee structure is essentially designed for high-balance accounts.
When the MFW can make sense:
- You have a $200,000+ TSP balance so the fixed fees are diluted to under 0.4%
- You want exposure to asset classes the core TSP doesn’t cover — emerging markets beyond the I Fund, REITs, commodities, sector funds, ESG funds, or actively managed alternatives
- You’re disciplined enough to trade rarely (each trade is $28.75)
When it doesn’t make sense:
- You’d use it to buy an S&P 500 index fund (the C Fund already covers this for free)
- Your balance is under $100,000
- You’d plan to do Roth in-plan conversions (MFW balances must be moved back to core funds first, which adds time and trade fees)
You can never have more than 25% of your total TSP balance in the MFW. If a strong year pushes you above 25%, you can’t add more until your balance rebalances back below the threshold. This cap by itself makes the MFW unsuitable as a primary investment vehicle.
9. Picking a mix that actually fits your career stage
The honest answer to "what allocation should I use?" depends on three things: how far you are from retirement, how secure your future pension is, and how much short-term volatility you can stomach without panicking.
Below is a defensible framework, not a recommendation. Adjust based on your specific situation.
| Stage | Age range | G/F | C | S | I | Equity total |
|---|---|---|---|---|---|---|
| Early-career | 22-40 | 5% | 60% | 15% | 20% | 95% |
| Mid-career | 41-55 | 15% | 55% | 12% | 18% | 85% |
| Pre-retirement | 55-62 | 30% | 45% | 10% | 15% | 70% |
| Early retirement | 62-72 | 25% | 50% | 10% | 15% | 75% |
| Late retirement | 72+ | 35% | 40% | 10% | 15% | 65% |
A few principles built into these numbers:
Equity exposure stays high even in retirement. A federal retiree with a 30-year horizon needs growth as much as any accumulator. The G/F percentage shown is a tactical cash bucket — typically 3-5 years of withdrawal needs in G Fund — not a strategic core.
The stock split is 60-65% C, 15-20% S, 15-20% I. This roughly matches global stock market capitalization while keeping a slight home-country bias most American investors are comfortable with.
The F Fund mostly stays at zero. As section 3 showed, the F Fund doesn’t add much over G Fund for most TSP participants. Some allocations include 2-5% F for diversification, but it’s rarely a high-impact move.
These allocations assume a participant with a FERS pension. Anyone without a pension — uniformed services members under BRS who separate before 20 years, or federal employees who took deferred retirement — should run a different calculation that puts more pressure on the TSP to replace lifetime income.
For TSP allocation during working years specifically, Federal Warrior covers Federal Warrior's career and pay coverage ↗ for accumulators at every GS level.
Frequently asked questions
Which TSP fund has the highest return?
Since its 1988 inception, the C Fund has the highest annualized return at 11.5%. Over more recent 10-year periods, the C Fund’s annualized 15.7% has also outperformed the S Fund (11.3%) and I Fund (10.4%). However, in any given year the leader changes — the I Fund led 2025 with returns over 32%, and was the top fund through Q1 2026 at +14.2% YTD. Long-run versus short-run leadership are different questions.
What’s the difference between the C Fund and the S Fund?
The C Fund holds the 500 largest U.S. companies (S&P 500). The S Fund holds essentially every other publicly traded U.S. company — about 2,500 small and mid-cap stocks. Owning both gives you full U.S. stock market exposure. The standard weighting is roughly 78% C and 22% S, matching their share of total U.S. market capitalization, though many participants simplify to 80/20.
Should I move my TSP into the Mutual Fund Window?
For most federal employees, no. The Mutual Fund Window costs at least $150 in annual fees plus $28.75 per trade, on top of the underlying mutual fund’s expense ratio. At balances under $100,000, those fixed fees create a fee drag of 0.4% or higher — significantly worse than the core TSP’s roughly 0.05%. Only about 0.13% of TSP participants use the MFW, almost all with high balances and a specific need for asset classes (emerging markets ex-I-Fund, REITs, commodities) the core TSP doesn’t cover.
Why are TSP fees so low compared to a 401(k)?
Three reasons: scale (the TSP manages roughly $1 trillion across 7.3 million participants — bigger than any single private 401(k) plan), structure (the FRTIB is a statutory agency operating at cost rather than for profit), and competitive sourcing (BlackRock and State Street manage the actual index tracking on negotiated institutional pricing). The result is total expense ratios of about 5 basis points, compared to typical private 401(k) plans charging 50-150 basis points.
What is the TSP I Fund actually invested in after the 2024 change?
The I Fund tracks the MSCI ACWI IMI ex-USA ex-China ex-Hong Kong Index — roughly 5,500 companies across 44 countries, including 21 developed markets (Japan, UK, Canada, Germany, France, Australia, etc.) and 23 emerging markets (India, Brazil, Korea, Taiwan, Mexico, and others). The 2024 update dramatically expanded the I Fund from its previous narrow EAFE benchmark, which only covered about 790 stocks in 21 developed markets and excluded Canada entirely. China and Hong Kong remain excluded due to congressional restrictions.
- TSP.gov G Fund (May 2026)
- TSP.gov C Fund (May 2026)
- TSP.gov S Fund (May 2026)
- TSP.gov F Fund (May 2026)
- TSP.gov I Fund (May 2026)
- TSPFolio fund data hub (May 13, 2026)
- NARFE Up-to-date Information — TSP performance through April 30, 2026
- FedSmith, "TSP Returns Slide in Early 2026" (April 2026)
- FedSmith, "TSP Returns Rocket Up In April" (May 2026)
- FedSmith, "Why Are So Few Federal Employees Using The TSP Mutual Fund Window?" (April 2026)
- FedTools, "TSP Mutual Fund Window 2026: Is It Worth the $132/Year?" (Feb 2026)
- TSP.gov Fund Information May 2026 (PDF)
- 5 CFR Part 1601 Subpart F — Mutual Fund Window