The Two I Funds (Old vs New)
I routinely get questions regarding the new I Fund and if it is better than the old I Fund. Presumably it would be or why switch, right? The old one wasn’t all that good. “Fair to middlin at best”, might be an apt description if you grew up near Barfield Financial’s current world headquarters. Let’s take a look at the two using hard numbers and data and limiting our opinions to see if we can determine if TSP made an improvement when they switched.
First of all, if you’re new to the TSP, you might be wondering what the heck I’m even talking about. A little backstory.
The original I Fund was born on May 1st, 2001. As with all other TSP Funds (besides G), the I Fund follows an underlying “index fund”, or a group of companies that are numerically tracked by a private company or system. In this case, it’s Morgan Stanley. The original I Fund tracked the performance of the index called MSCI EAFE. That is a pretty well-known international index. And the TSP used it until October of 2024.
On October 30, 2024, the TSP stopped using the MSCI EAFE and switched to the MSCI ACWI IMI ex USA ex China ex Hong Kong index. Yes, that is the actual name of it. The thinking behind this switch was that it would be better than the old one that was limited in the places of the world it could invest. The old one only tracked about 800 companies over 21 developed countries.
The new I Fund index tracks about 5,000 companies over 21 developed countries and 23 emerging market countries. So it’s far more diverse and broad. People with a desire to invest internationally were asking for this and they got it.
This causes some problems though when you’re trying to track the data. Particularly long term returns. Because the new I Fund has only been around for 2 years, and the old I Fund we stopped using 2 years ago. But when you go onto TSP.gov, you’ll see I Fund data going back to 2001. That’s not the current I Fund though. That old data is from…well, the old I Fund. You see? First I Fund was 2001 to 2024, then we switched to the new I Fund from 2024 until now.
(If you want to split hairs, it’s technically the same I Fund, but the fund itself is now tracking a different index than it used to track if that makes sense. For most of us, from a practical standpoint, that is a distinction without a difference. It means nothing, but there are plenty of finance bros and bras here that like to be absolutely 100% as specific as possible so I want to qualify that.)
So here’s the question:
Are the past results of the old I Fund index useless now that we have the new I Fund index? Or are the results of the old I Fund index similar (and therefore useful to us) enough to compare the new I Fund index?
Well, it turns out the results of the old I Fund index and the new I Fund index are very, very similar. The new I Fund index is not substantially better than the old one. But it is technically better, mathematically, by a little bit. Which means that the data on the I Fund as a whole, is pretty accurate going back years, even though the underlying index changed in 2024.
Read on for some hard data.
I Fund Index Two (the newer one) was created on June 7, 2023. So we don’t even have real data before then. It didn’t exist. But what the TSP has done is “backtested” the new I Fund back to 2003 (Actually TSP hired Aon Investments USA, Inc to do the actual math). Backtested simply means “If this thing had existed in the past, this is what it would have made each year.” It can be thought of as a theoretical, or reconstructed, return. For those of you that use certain TSP allocation strategies, backtesting is popular and common. They might say something like, “If our current strategy had been around for the last 25 years, these are the returns you would have made.”
No one has actually made those returns during backtesting. Because the strategy or the index or the investment didn’t actually exist. But the calculations are extrapolated backwards to reconstruct it artificially. That’s backtesting.
Here’s a simplified (and absurd) example of backtesting. Someone notices that every day Jim comes into the office wearing a red shirt, the market goes up that day. The office gets a little excited and does some backtesting. They go review security footage over the last 10 years and write down every day that Jim wore a red shirt to work. Then they take all those dates and compare them to the stock market performances for those same days and, wouldn’t you know it—lo and behold, 79% of the time Jim wears a red shirt to work, the market goes up. A spreadsheet is created with this new “foolproof” TSP strategy that shows how someone can successfully time the market by just figuring out when Jim is going to wear a red shirt. And, hey, “THE DATA PROVES IT! WE BACKTESTED THE DATA! LIKE ACTUAL SCIENTISTS!” And boom…another TSP Facebook page (and resulting cult) is born. The Red Shirt Strategy has returned 3,000% more than the SP500. You’re a fool not to be following it. Haven’t you heard?!?
Well, has it really? The Red Shirt Strategy didn’t actually exist 10 years ago so not a single human being has actually made 3,000% more than the SP500 using this strategy. Will it continue to beat the market 79% of the time in the future? It certainly doesn’t seem logical, but hey, who knows? There’s crazier things on Facebook.
So backtesting can be useful, but please be aware of its limitations.
The result is that we have a reported backtested return of the new I Fund going back to the full year starting in 2003. We have actual TSP results from the old I Fund index going back that far already because it did in fact, exist. So it’s a simple matter of putting each I Fund’s index return down for each year side by side, and calculating the total return for each. This is often called the Compound Annual Growth Rate (CAGR). Aka what the investment has averaged compounded each year over that total period of time.
This shows the actual growth of an investment over time, calculating each year’s loss or gain, and then averaging it out to show what it’s average annual return was. Here are the results:
Old I Fund Index (MSCI EAFE)
2003-2025 CAGR: 7.77%
New I Fund Index (MSCI ACWI IMI ex USA ex China ex Hong Kong)
2003-2025 CAGR: 8.44%
This data is pulled directly from the TSP’s November 2023 Benchmark Study, and then for the couple of years after the report, we can pull the returns from Morgan Stanley itself who reports the returns. (Links to documents below).
The other data points comparing the old and new I Fund indexes (or indices), are still the same if you want to compare things other than just the return, like fees and so forth:
I Fund remains the riskiest TSP fund
I Fund fees remain higher than C Fund and lower than S Fund
The I Fund indexes (both old and new) have the greatest number of down years since 2003, compared to G, F, C and S