Okay, everyone read the string that MrSpendy linked. If this thread is useful for you at all.
Hooooooooly shit, what an emotional roller coaster! So much detailed content in that string. It took me the better part of 24 hours to read, reread, digest, and interpret, and that's just for the stuff I could understand. Your input was really valuable to that string (and also therefore this one), MrSpendy. Thank you. Radagast and L.A.S also had some really great input.
Here's the really short answer, to whomever encounters this string, as far as I'm concerned: it doesn't really matter. If - for any of the
good reasons that exist (
not any of the bad ones!) - you want to diversify your portfolio to capture lower risk / lower volatility / lower return holdings, look around to find some good options, but don't look too hard because you're wasting your time. Getting good CD rates and laddering is viable, keeping a secondary EF in HY cash is fine, buying i-bonds isn't crazy, indexing to a sensible bond fund makes sense. Also long as these holdings don't represent a giant portion of your portfolio, it won't matter terribly, unless you're really cutting it close on the funds you need, about the return because there isn't a best option and the spread of returns is a pretty narrow band. Most mixes of these holdings will provide about the same return, basically. If you're pinching pennies so hard that this portion of your portfolio really matters, you probably just need to work a couple more years.
The juiciest stuff that had the most impact on me is here:
From L.A.S:
Generally speaking bonds move inversely to the direction of prevailing interest rates. So if the prevailing yield for a bond in the market moves up, then the price of the bonds will go down. It is possible to loose money in investment grade bonds.
The rule of thumb for bond funds is that for rapid movements in market yield, the price of the fund will move in an amount of (-1) x rate change x (weighted average duration in years) of the bonds in the fund. So if your bond fund has a duration of 7.2 years, and interest rates rise 2% over a short period of time -- say six months or so -- with respect to intermediate to long term bonds, then the investor will probably have suffered approximately a 14-15% capital loss in their holding. If interest rates drop 2% then the investor would have a gain of 14-15% or so. However, the exact amount would of course be set by the market. This is interest rate sensitivity risk.
But, over long periods of time the volatility in bonds will be smoothed out just like with stocks. In the case of drops, this is due to the fund having a higher yield as well as interest distributions being reinvested at that higher rate.
People have about as much luck predicting the direction of interest rates as they do predicting what the stock market is going to does next.
One can protect themselves from interest rate sensitivity risk by buying shorter term bonds.
MrSpendy gold:
My opinion is that the bond market should be "timed" much more so than the equity market. one should weigh the potential risks versus potential rewards from investing in the bond market more so than the equity market. this is because the upward bound of return from an individual bonds (and to a lesser extent the bond market) is much more knowable and that bond market has competitors for your investment dollars (CD's and savings bonds) that are very legitimate alternatives to the Vanguard Total Bond market index.
For example, I don't think it is hopeless market timing or a bad thing to say "the bond index yields 2.3% with duration of 6, I can buy a CD or saving bond with a similar yield and liquidity without credit or duration risk / reward, I'll do that instead". it is market timing in that you will miss out if rates go down and the bond market experiences capital appreciation. your return will vary from the "passive alternative".
if the role of bonds in the portfolio is stability and you can get more stability with the same hold to maturity type of return, why not?
there is a sophisticated retort to this line of reasoning involving bonds' correlation with stocks and the diversification benefits, but I'd counter that CD's / savings bonds correlation of 0 to everything but a collapse of the FDIC / government also works and that historical correlation of bonds / equities (particularly during the last crisis where bonds shot up when equities went down) may not repeat.
in short, I think it's okay to decide when to not take duration or credit risk for the safe part of the portfolio and to weight the risks versus the reward.
it's not the same as saying "I think this company will outperform the stock market", it's saying making 2% isn't worth it if I can lose 7% if rates increase by 1% or 2.7% isn't worth it to me if I can lose 18% if rates go up by 1% (in the case of the 30 year treasury).
to answer the question about the long term bond fund. I would say that you should buy it if you want to make 3.5% yield + - the return from duration+a little credit risk. if that has a role in your portfolio, then buy it. if not, then don't buy it. I personally think you should have a very good reason for buying bonds with the most risk and reward in terms of duration. You don't appear to have one so I'd advise against it.
From Radagast:
Yes, at your stage you should have bonds. When you are halfway/within 5 years of FIRE start gliding towards 10-25% bonds to minimize sequence of returns risk (up to 40% if you are scared of stocks or planning a reverse glide path). Move your cash to bonds ASAP, because cash is even less useful, unless you need it tomorrow.
VBTLX is fine. If you want to go yield chasing for CD's that is fine too. Don't over think it, because the rewards for successfully timing the bond market are very tiny.
VSBSX has a yield of 1.28%. Say inflation is 2%. You lost money. Good job. What where you trying to do again?
Rates rise, VSBSX loses 2% and pays back 2%, VBTLX loses 7% and pays back 3%. In 2.5 years VBLTX will come out ahead. In fact it will come out ahead after 3 years about 95% of the time (I made that up, not looked it up).
Say VBLTX lost 5% more than VSBSX did because of a 1% increase in yield, and bonds were 20% of your portfolio. Well, you suffered a huge 1% loss over the course of a year because of your choice of bond funds. Big whoop, your stock funds lost more than that several times this week. That is why timing the bond market is virtually pointless, and differences between bond funds are mostly not important. I see no reason to use VSBSX when the most likely result is having less money.
It is easy to get into the weeds thinking about bonds with virtually no meaningful difference in outcome.
"I use intermediate term bond index because I don't like the convexity of MBS" OK
"I only use 3-7 year treasury bonds because academic research shows term and credit risk don't pay" Fine (what is inflation?)
"I listen to Jack Bogle and add corporate bonds, because total bond doesn't have enough" whatever
"I use a barbell of short and long term treasury bonds" carry on
"I have 12 CD's from 7 online credit unions because that has given me an extra .5% return (we're talking about 10 bps per year on my portfolio people!!!) without the interest rate risk compared to bonds, plus the early withdrawal penalty is less than interest rate risk 12% of the time!" Ummm alright
"I divide my bond allocation among total bond, long duration zero coupon treasury bonds, high yield tax exempt bonds, series I savings bonds, hedged international bonds, TIPS, plus some CD's at various banks" Ok, wow
None of the above really matter on 10-20% of the total. If you start getting into long term bonds it can matter a little.
There was some stuff I didn't really understand, too.
Here's a question: If I'm willing to get 3.5% in highly secure returns to reduce volatility in our total holdings, are EE bonds reasonable? I'm thinking to execute on a mix of EE bonds, CDs, and HY savings where I want true liquidity. Let me also just point out that there are high income taxes for our bracket in our state and we're planning on staying here, so not paying the state taxes on EE bonds is also appealing. I assume that taxes are due in the state of residence at time of redemption.
For those not doing the math or not familiar with the product, you can buy a bond (EE Bond) that is currently guaranteed (by the US government) to double in value in 20 years. The intermediate returns are close to zero, but then they up the return at the 20 year mark so the bond has doubled in value. If you do the math on 10k (which is the max per person (by SSN) per year) turning into 20k over 20 years, that's a 3.5% compounded APY. Doing that while also being on the lookout for good CDs and laddering those seems at least as good as parking dough in a bond index. Also, if exclusively using CDs and EE bonds, there is no risk of those investments losing value (at my expected contribution levels, not hundreds of millions) other than opportunity cost.
Correct me where I'm wrong, please! And again, thanks for that link!
edited to comment on high state taxes and how great it would be to not pay them on EE bond redemption.
