TSP - Treasury, Bond and Stock Funds

[Purron]

I've been going over the recently revamped TSP site (www.tsp.gov). I've got most of my funds in G and have been thinking of moving some into other funds to diversify. I'm also considering moving some IRA funds I have in banks and credit unions into the TSP because of the super low rates out there. I know I tend to be conservative (to a fault) and worry about missing out on some gains.

Here's a description of the 3 most popular funds from the TSP site:

G - The G Fund invests exclusively in a nonmarketable short-term U.S. Treasury security that is specially issued to the TSP. The earnings consist entirely of interest income on the security.

F - The F Fund invests in a bond index fund that tracks the Barclays Capital U.S. Aggregate Bond Index. This broad index includes U.S. Government, mortgage-backed, corporate, and foreign government (issued in the U.S.) sectors of the U.S. bond market. The earnings consist of interest income on the securities and gains (or losses) in the value of the securities.

C - The C Fund invests in a stock index fund that fully replicates the Standard and Poor's 500 (S&P 500) Index. The earnings consist primarily of dividend income and gains (or losses) in the price of stocks.

Here are the returns since inception and for the past 10 years:

G - 6.15% since inception date of 4/1/87, 4.62% last 10 years
F - 7.10% since inception date of 1/29/88, 6.39% last 10 years
C - 9.31% since inception date of 1/29/88, (0.94%) last 10 years

I've always thought stocks and bonds yielded much more than treasuries in the long term and this is the price I’m paying for the lower risk G fund. In looking at the long term returns, however, I was surprised to see a much smaller difference than I anticipated.

I'm now questioning moving anything out of the G fund. Why give up the safety of the G fund for a relatively small difference? I realize over time the difference in returns would be pretty significant, but I'm not getting any younger and don't know if the risk is worth it. What am I missing here?

BTW, the annualized rate for August 2010 in the G fund was 2.64% - low, but not bad these days. For the last 12 months, the G averaged 3.10%.

 

[W2R]

In accumulation phase, I made a killing in the TSP's S fund (small cap) and I fund (international) and did not mess around with G Fund much during market upturns.

But now that I am retired, I am sticking to 100% G Fund because for me, the TSP is like a second pension. I am withdrawing in equal monthly payments and I know I can depend on them. I am also somewhat (at least slightly?) confident that G Fund will keep up with inflation in the long run.

I have plenty of room in my taxable account for all the equities I need for my overall 45:55 (stocks:bonds) asset allocation, plus more bonds. So, my TSP is just part of my bond holdings. I don't expect returns from bonds to equal returns from equities in the long run (though who knows, but if I did then I wouldn't hold any equities at all).

With all the talk about moving into 100% cash, or all bonds and ditching equities, and with the crash of 2008-2009, I am beginning to think that this is probably an excellent time to buy equities. Like Rothschild said, roughly, buy when there is blood in the streets.

On the other hand, I don't really have to so I am sticking with my AA and rebalancing plan.

 

[Purron]

With all the talk about moving into 100% cash, or all bonds and ditching equities, and with the crash of 2008-2009, I am beginning to think that this is probably an excellent time to buy equities. Like Rothschild said, roughly, buy when there is blood in the streets.

Yes, it might be a good time to buy. But I'm no good at reading the tea leaves and even worse at trying to time the market. My point is, in the long run, the C fund (stocks) hasn't outperformed the G fund (treasuries) by near as much as I thought. I read a lot of stuff about people missing out on big upturns in the stock market, but when you average in the down times, how good is it really? Good enough to offset the risk? Good for an aging boomer who may not have time to wait for the long-term gains when you average a portfolio's performance out over a couple of decades?

BTW, I'm not considering making a drastic move here. Just looking at moving some more into the TSP and trying to figure out where to put it. Bottom line, no one knows where we'll be 10 or 20 years from now. I'm not a doom and gloomer and believe things will improve. I also would guess we're in for a bumpy ride for the next few years or maybe a painfully slow return to growth like the Japanese have endured.

 

[W2R]

You are SO RIGHT about market timing and tea leaves. I just never seem to do well with it, so I think about it but try not to act on market timing impulses. I probably shouldn't have put that paragraph in my last post because it confuses the issue.

I think that asset allocation should be determined for your entire portfolio regardless of how much of that portfolio is in the TSP. So, you have, say, a certain percentage in equity funds and a certain percentage in bond funds. As you point out, your percentage in bond funds should usually increase as you get close to retirement.

Then personally I prefer the G fund to other bond funds that I can get outside of the TSP, such as the Vanguard Total Bond Market Fund. For me, it does no good to compare the G fund with any equity fund because that would be apples to oranges, and I already know that I need a certain percentage of bond funds (of which this is a part).

The C fund is designed to simulate the S&P 500. So, it is easy to compare that with equity funds outside the TSP.

As you know conventional wisdom says that bond funds are better off tax sheltered, such as within the TSP.

If all your investments are within the TSP (I have no idea, but mine are not), I would determine percentages depending on your asset allocation percentages. Volatility is a concern as well as overall yield. The G Fund was as stable as a rock during 2008-2009 and that is helpful in keeping one from selling equities in a panic when they plummet.

 

[Independent]

Here are the returns since inception and for the past 10 years:

G - 6.15% since inception date of 4/1/87, 4.62% last 10 years
F - 7.10% since inception date of 1/29/88, 6.39% last 10 years
C - 9.31% since inception date of 1/29/88, (0.94%) last 10 years

I've always thought stocks and bonds yielded much more than treasuries in the long term and this is the price I’m paying for the lower risk G fund. In looking at the long term returns, however, I was surprised to see a much smaller difference than I anticipated.

The big difference takes some compounding before it appears. How about assuming that inflation was 3%? Then the G fund doubled your real spending power in 23 years, the C nearly quadrupled it. That seems like a big difference to me, and 23 years seems like a reasonable period from saving to spending.

OTOH, if I'm retiring today and know I'll be spending a lot of the money in the next 10 years, then I'd prefer the safety of the G fund.

 

[freebird5825]

Since Dec 12, 1988 (SCD), I proportioned my TSP contributions at 60/40.

My algorithm :
60% in C and 40% spread equally between the G and F funds.

When the S and I funds came along, I changed the 60% to be spread equally across C,S,I and 40% spread equally as before in the G and F funds.

Every time we were allowed to increase the percentage going in, I jumped right on that. When the dollar limit was increased by the IRS, I was on top of that.

When the L2020 fund came out, I transferred 50% of all existing funds into that, left 50% in the 60/40 AA mode. All new money went in as 60/40 to C,S,I and G,F as usual. I rebalanced quarterly if it made sense, i.e sent unrealized gains to the L2020 fund with intrafund transfers. No tax consequences while still w*rking.

Why am I telling you this? Sometimes it doesn't hurt to "set aside" half of your accumulation in one of the L funds on auto-pilot, and put "new" money into individual funds as you feel comfortable.

I left govt service in 1Q07 (lucky as hell!) with a hefty sum (took the SEPP option) accumulated over 18 years 3 months.

 

[jazz4cash]

I am fascinated with the G fund (ever since W2R first mentioned it years ago). My spouse is a Fed since '04, so I am still learning the wrinkles of TSP.

It occurs to me that G fund is not really a bond fund since "The earnings consist entirely of interest income on the security" there is no interest rate risk, so the fund should be considered a cash allocation. What am I missing?

 

[Bimmerbill]

The G fund is almost a money market fund that pays extremely well. The bond duration is almost 0, just like cash. I think it used to be 1 to 3 days in duration, if I remember correctly.

 

[clifp]

The big difference takes some compounding before it appears. How about assuming that inflation was 3%? Then the G fund doubled your real spending power in 23 years, the C nearly quadrupled it. That seems like a big difference to me, and 23 years seems like a reasonable period from saving to spending.

OTOH, if I'm retiring today and know I'll be spending a lot of the money in the next 10 years, then I'd prefer the safety of the G fund.

I agree considering what horrible decade 2000 was for stocks, I am impressed there is that much different between what is inarguable one of the best bond funds the G fund, vs a low cost index stock fund.

It would be interesting to see how much a "typical" government worker who maxed out their TSP and received average raises would have been able to save using different AA studies. It seems look a good academic study, to see what is theoretically possible for a defined contribution plan.

It sounds like freebird did a good job.

 

[ziggy29]

The problem is that the "last 10 years" -- indeed, the last 28 years with a few rough spots -- have come in an extreme bull market for bonds while stocks flatlined at best. Anything is possible, but given their current relative valuations I have a hard time believing we'll see bonds continue a bull run for another decade and outperform stocks through 2020. Could be wrong...