testing the Ls

[Bubba]

I'm just curious if anyone's been testing using an L instead of the G or F (which months, which type, etc.). Particularly I'm interested in the Linc and the L2060 or L2065. Others ok too. The point is that they might be interesting alternatives instead of using the G or F funds.

[Aitrus]

I don't know of anybody that has. I played around with the idea a bit when they first came out, but I didn't for several reasons.

1 - There weren't any historical datapoints I could pull from. It would take a long time for the dataset to populate. Doing so now would take a lot of work, and I haven't done any serious spreadsheet updates since 2021 or so.

2 - The L Funds aren't static. Meaning, that if one made a strategy using L2030 and did so in 2020, then in 2030 that strategy wouldn't be representative of what L2030 looked like anymore. This is because what L2030 looked like in 2020 is very different than what it will look like in 2030, when it will match the L-Income Fund. Somebody would need to constantly update their data each month as TSP updates the Fund composition, and that's a lot of work with little benefit. Probably easier now with the advent of tools like PowerBi and various AI assistance approaches, but it's still a lot of work.

About the only L-Fund that this approach might work with is L-Income, because it's static over the long-term (except in 2018, when the L-Income Fund started slowing increasing it's C-S-I stock allocation from 20% to 30% over a period of 10 years).

If you want to replicate a permanent-style L2060 or L2065, just do a strategy that allocates 50% C, 15% S, and 35% I. This will closely mirror what L2065 looked like when it first came out in 2020.

MJ added an ability to do something like this to the TSPCalc site by including three split options to the Seasonal Calculator: 25% C/75% S, 50C/50S, and 75C/25S. I think this is a close as anybody has come to doing what you suggest.

[Bubba]

I don't know of anybody that has. I played around with the idea a bit when they first came out, but I didn't for several reasons.

1 - There weren't any historical datapoints I could pull from. It would take a long time for the dataset to populate. Doing so now would take a lot of work, and I haven't done any serious spreadsheet updates since 2021 or so.

2 - The L Funds aren't static. Meaning, that if one made a strategy using L2030 and did so in 2020, then in 2030 that strategy wouldn't be representative of what L2030 looked like anymore. This is because what L2030 looked like in 2020 is very different than what it will look like in 2030, when it will match the L-Income Fund. Somebody would need to constantly update their data each month as TSP updates the Fund composition, and that's a lot of work with little benefit. Probably easier now with the advent of tools like PowerBi and various AI assistance approaches, but it's still a lot of work.

About the only L-Fund that this approach might work with is L-Income, because it's static over the long-term (except in 2018, when the L-Income Fund started slowing increasing it's C-S-I stock allocation from 20% to 30% over a period of 10 years).

If you want to replicate a permanent-style L2060 or L2065, just do a strategy that allocates 50% C, 15% S, and 35% I. This will closely mirror what L2065 looked like when it first came out in 2020.

MJ added an ability to do something like this to the TSPCalc site by including three split options to the Seasonal Calculator: 25% C/75% S, 50C/50S, and 75C/25S. I think this is a close as anybody has come to doing what you suggest.

Thanks! When I was playing with the data (outside the TSP), I played with AOK, but I found that ETF not as helpful...still indicative of a what a return might look like.