Yield Curves-Treasury, CD, AA/AAA muni

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Sun Apr 22, 2018 9:27 pm

Here's a chart of the yield curves I've been using for my fixed-income investment decisions lately. This is from Friday, or whatever yields Fidelity is showing right now. I pull the yields from the Fidelity Bonds & CDs yields webpage into a Google Sheets spreadsheet using IMPORTHTML to construct this.

Image

Before some observations, here are some caveats and explanations.

Treasury yields typically are for large quantities, like 100 ($100,000 face value) or more, and may be for maturities that are as much as one month more than indicated. So the yields I get probably are a bit lower.

TEY stands for Taxable Equivalent Yield, and are the TEYs for my marginal federal and state tax rates of 27% and 8%, not itemizing deductions. I convert anything that has a federal or state tax exemption to TEY so I can compare apples to apples for investments in taxable accounts. Note for example that Treasury TEY is higher than Treasury, and that Treasuries could have higher TEYs than CDs in taxable accounts, but lower yields in tax-advantaged. This is because of the state tax exemption and my 8% marginal state tax rate.

CD yields are for new-issue CDs. I usually can find slightly higher yields for same maturities in secondary market.

There are no commissions for Treasuries or new-issue CDs, so these yields do not need to be adjusted for commissions. But remember that Treasury yields probably are higher than what I would actually pay for smaller quantities at maturities closer to those indicated, but the yields are good enough for initial decision making.

There is a $1 per bond commission at Fidelity for secondary munis (so 0.1%), so muni yields are higher than what I'd actually get. Also, the yields are for munis that I might not want to buy, either because they have undesirable call features, or are in municipalities that I do not want to buy (or buy more of). Still, this gives me a rough idea of the relative attractiveness of munis in taxable accounts.

For now I'll just observe that what looks best to me depends on whether it's in a tax-advantaged or taxable account, and the particular term to maturity I'm looking at. I am not dogmatic about what I'll buy--I have been buying Treasuries, CDs and munis.

Kevin
Last edited by Kevin M on Mon Apr 23, 2018 10:43 am, edited 1 time in total.
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Mon Apr 23, 2018 1:23 am

A few more clarifications and observations.

For 0-year yield in tax-advantaged accounts, I use Prime money market yield, currently 1.79%. So this is the baseline for Treasury and CD curves, as those are what I'd look at in tax-advantaged.

For 0-year yield in taxable I use my TEY for California muni money market fund, currently 2.22%.

Of course these MM funds have more credit risk than a Treasury MM fund, but IMO it's so small as to be negligible, and those are the funds I use if possible (at Fidelity, the MM funds are 10 basis points or more lower yield).

Note that CD and Treasury TEY yield curve slopes are negative out to 3 months, so no point in buying any of these at 3-month maturity right now. Barely enough justification to buy them at 6 months. At 1-year these look better.

AAA muni curve looks very steep out to 6 months, but commission will cut into this a lot, so not as good as it looks. Commission hits more at shorter maturities, since it's allocated over less time.

AA muni curve looks really steep to 1-year, but you need to drill down and evaluate the muni that this quote is for to see if it's something you're comfortable with and can buy. This particular quote is for minimum quantity 30 ($30,000 face value) of an AA- MI GO muni. For my portfolio, I won't put this much into one municipality (I tend to buy quantity 5 or 10), but otherwise I would consider it. Chances of a default of an AA- muni in one year are 0% historically, but the high yield tells us something about the market's perception of risk, and I don't ignore that. Next highest yields at this maturity are about 2%, compared to 2.3% for this one (before adjusting for tax exemption).

If I wanted to stick with AAA munis, I might as well go with Treasuries out to 2-year maturity, as the TEYs are about the same. At 3-year maturity I'd want to take a closer look at the AAA munis. But I will consider AA munis at 2-year maturity, so I've tended to buy more of those (and yields might have changed some since my purchases). Also, this chart doesn't show my TEY for CA AA and AAA munis, and sometimes those will be a better deal for me as a CA resident, but often not.

Treasury TEY beats CDs in taxable out to five years, but CDs take the lead over Treasuries in tax-advantaged at 2-year maturity.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

ivk5
Posts: 405
Joined: Thu Sep 22, 2016 9:05 am

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by ivk5 » Mon Apr 23, 2018 3:02 am

Kevin - this is terrific - thanks as always for walking us through your FI thought process.

ofckrupke
Posts: 531
Joined: Mon Jan 10, 2011 2:26 pm

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by ofckrupke » Mon Apr 23, 2018 10:16 am

Kevin M wrote:
Sun Apr 22, 2018 9:27 pm
TEY stands for Taxable Equivalent Yield, and are the TEYs for my marginal federal and state tax rates of 28% and 8%, not itemizing deductions.
Is this a typo? Previously you posted that your TCJA marginal rate for interest income is 12+15=27% (LTCG/QDIV straddles the start of the 15% rate).

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Mon Apr 23, 2018 10:44 am

ofckrupke wrote:
Mon Apr 23, 2018 10:16 am
Kevin M wrote:
Sun Apr 22, 2018 9:27 pm
TEY stands for Taxable Equivalent Yield, and are the TEYs for my marginal federal and state tax rates of 28% and 8%, not itemizing deductions.
Is this a typo? Previously you posted that your TCJA marginal rate for interest income is 12+15=27% (LTCG/QDIV straddles the start of the 15% rate).
Yes, typo. Fixed. Thanks for the catch!

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

invstar
Posts: 95
Joined: Tue Apr 25, 2017 1:20 pm

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by invstar » Mon Apr 23, 2018 11:08 am

:thumbsup
Kevin - Thanks for all the great info. I am learning a lot from your posts..
Kevin M wrote:
Mon Apr 23, 2018 1:23 am
A few more clarifications and observations.

For 0-year yield in tax-advantaged accounts, I use Prime money market yield, currently 1.79%. So this is the baseline for Treasury and CD curves, as those are what I'd look at in tax-advantaged.

For 0-year yield in taxable I use my TEY for California muni money market fund, currently 2.22%.

Of course these MM funds have more credit risk than a Treasury MM fund, but IMO it's so small as to be negligible, and those are the funds I use if possible (at Fidelity, the MM funds are 10 basis points or more lower yield).

Note that CD and Treasury TEY yield curve slopes are negative out to 3 months, so no point in buying any of these at 3-month maturity right now. Barely enough justification to buy them at 6 months. At 1-year these look better.

AAA muni curve looks very steep out to 6 months, but commission will cut into this a lot, so not as good as it looks. Commission hits more at shorter maturities, since it's allocated over less time.

AA muni curve looks really steep to 1-year, but you need to drill down and evaluate the muni that this quote is for to see if it's something you're comfortable with and can buy. This particular quote is for minimum quantity 30 ($30,000 face value) of an AA- MI GO muni. For my portfolio, I won't put this much into one municipality (I tend to buy quantity 5 or 10), but otherwise I would consider it. Chances of a default of an AA- muni in one year are 0% historically, but the high yield tells us something about the market's perception of risk, and I don't ignore that. Next highest yields at this maturity are about 2%, compared to 2.3% for this one (before adjusting for tax exemption).

If I wanted to stick with AAA munis, I might as well go with Treasuries out to 2-year maturity, as the TEYs are about the same. At 3-year maturity I'd want to take a closer look at the AAA munis. But I will consider AA munis at 2-year maturity, so I've tended to buy more of those (and yields might have changed some since my purchases). Also, this chart doesn't show my TEY for CA AA and AAA munis, and sometimes those will be a better deal for me as a CA resident, but often not.

Treasury TEY beats CDs in taxable out to five years, but CDs take the lead over Treasuries in tax-advantaged at 2-year maturity.

Kevin

User avatar
Doc
Posts: 8614
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Doc » Mon Apr 23, 2018 11:54 am

Kevin M wrote:
Mon Apr 23, 2018 1:23 am
For 0-year yield in tax-advantaged accounts, I use Prime money market yield, currently 1.79%. So this is the baseline for Treasury and CD curves, as those are what I'd look at in tax-advantaged.
What is the rationale for using a MM yield for 0-year. Why not a 4 wk T-bill or just assume it as cash and use zero.

It would have an effect on the negative slope:
Kevin M wrote:
Mon Apr 23, 2018 1:23 am
Note that CD and Treasury TEY yield curve slopes are negative out to 3 months, so no point in buying any of these at 3-month maturity right now. Barely enough justification to buy them at 6 months. At 1-year these look better.
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

workingatit
Posts: 89
Joined: Fri Apr 20, 2007 8:16 pm

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by workingatit » Mon Apr 23, 2018 1:57 pm

what will be the impact of the multiple expected rate increases this and next year on this data? iow, if yields are expected to go up, why would one want to buy and lock in these current rates?

and Kevin, this is great information, thank you

aristotelian
Posts: 4686
Joined: Wed Jan 11, 2017 8:05 pm

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by aristotelian » Mon Apr 23, 2018 2:26 pm

workingatit wrote:
Mon Apr 23, 2018 1:57 pm
what will be the impact of the multiple expected rate increases this and next year on this data? iow, if yields are expected to go up, why would one want to buy and lock in these current rates?

and Kevin, this is great information, thank you
All other things being equal, you would expect all the curves to move up, while keeping the same shape. Of course, all other things are not always equal.

The interest rates controlled by the Fed are only one factor determining interest rates as a whole. If there was a stock market crash tomorrow, you might see Treasury rates go back to zero regardless of what the Fed does. If you thought the stock market was going to crash tomorrow, you would want to lock in these rates, even knowing that the Fed might increase its rates.

misterno
Posts: 478
Joined: Sat Mar 08, 2014 7:55 am

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by misterno » Mon Apr 23, 2018 2:28 pm

I am just surprised to see Corporate BBB rated bonds paying 2.29% maturing in August

What are the chances of being bankrupt in 3 months anyway?

ThrustVectoring
Posts: 498
Joined: Wed Jul 12, 2017 2:51 pm

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by ThrustVectoring » Mon Apr 23, 2018 4:12 pm

Have you looked into the after-tax return of treasury futures? Gains and losses get treated as 60% long-term / 40% short-term capital gains, which is probably advantageous.

User avatar
Artsdoctor
Posts: 3504
Joined: Thu Jun 28, 2012 3:09 pm
Location: Los Angeles, CA

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Artsdoctor » Mon Apr 23, 2018 4:34 pm

Baird compiles data that you might be interested in if you haven't already looked at it; there is an analysis of treasuries and municipals which can be helpful. They publish their synopsis every other week.

http://content.rwbaird.com/RWB/Content/ ... entary.pdf

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Tue Apr 24, 2018 4:39 pm

Doc wrote:
Mon Apr 23, 2018 11:54 am
Kevin M wrote:
Mon Apr 23, 2018 1:23 am
For 0-year yield in tax-advantaged accounts, I use Prime money market yield, currently 1.79%. So this is the baseline for Treasury and CD curves, as those are what I'd look at in tax-advantaged.
What is the rationale for using a MM yield for 0-year. Why not a 4 wk T-bill or just assume it as cash and use zero.
I think I explained that already. Basically I'm comfortable with the tiny risk of Prime MM in taxable and CA Muni MM in IRA, so that's where my cash in a brokerage account is going to be parked. If I were not comfortable with it, I'd probably use the SEC yield of Treasury MM. I'm certainly not going to hold cash at 0% when I can get 1.8% in Prime MM and 2.25% TEY in CA muni MM (except for what I actually plan to spend in cash in the next short period of time).

Of course you can use anything that makes sense to you.
It would have an effect on the negative slope:
Of course. This are my personal yield curves for my personal decision making. A generic yield curve would use the technically appropriate 0-year yield, or for example maybe like you say, the 4-week Treasury yield as the baseline for Treasuries. My TEY for 0-year muni is not going to be the same as someone with different marginal tax rates and/or living in a different state.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Tue Apr 24, 2018 4:51 pm

workingatit wrote:
Mon Apr 23, 2018 1:57 pm
what will be the impact of the multiple expected rate increases this and next year on this data? iow, if yields are expected to go up, why would one want to buy and lock in these current rates?

and Kevin, this is great information, thank you
You're thinking of the Federal Funds Rate (FFR), which is the shortest-term rate you can have (overnight). Yields at longer maturities are not necessarily impacted by changes in the FFR. I explained this in another thread recently, and showed a chart that shows that the 1-month Treasury tends to anticipate and track changes in FFR, but the 10-year Treasury yield, for example, does not have any definite relationship to changes in FFR.

Yield change expectations should affect the yield curve, in theory, with a positive yield curve reflecting expectations that shorter-term yields will increase in a way that you'd earn the same by rolling shorter-term securities as buying a longer-term security. The theory doesn't seem to be well supported by empirical evidence, or so I've read.

Other theories include liquidity-preference, which is just an academic way of saying that longer-term yields will be higher than based only on expectations due to the uncertainty of expectations being met. In other words, there is a yield premium for taking term risk.

The bottom line is that there are no good rate forecasters (with the FFR perhaps being somewhat of an exception as rates go), so no one really knows that a particular yield will change in a particular way over any particular time period.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Tue Apr 24, 2018 4:53 pm

misterno wrote:
Mon Apr 23, 2018 2:28 pm
I am just surprised to see Corporate BBB rated bonds paying 2.29% maturing in August

What are the chances of being bankrupt in 3 months anyway?
Good question. One thing is that commission really cuts into yields at shorter maturities. Rerun your yield calculation with commission, and that it may not look nearly as attractive as it seems.

I haven't paid much attention at all to corporates, but maybe I should.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Doc
Posts: 8614
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Doc » Tue Apr 24, 2018 4:59 pm

Kevin M wrote:
Tue Apr 24, 2018 4:39 pm
I think I explained that already. Basically I'm comfortable with the tiny risk of Prime MM in taxable and CA Muni MM in IRA, so that's where my cash in a brokerage account is going to be parked. If I were not comfortable with it, I'd probably use the SEC yield of Treasury MM.
I understand that. My question was not about what funds you used but why you used the MM funds as zero duration instruments. I don't buy MM instruments so I don't pay a lot of attention to them but you made me look. CA Muni has an average maturity of 29 days. VG Federal and Prime MM funds have a duration of 50 days. If you use the returns (SEC yield) from these MM instruments but used the actual maturity instead of zero duration would you still have the negative slope at the short end of your curve?
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Tue Apr 24, 2018 5:44 pm

Doc wrote:
Tue Apr 24, 2018 4:59 pm
Kevin M wrote:
Tue Apr 24, 2018 4:39 pm
I think I explained that already. Basically I'm comfortable with the tiny risk of Prime MM in taxable and CA Muni MM in IRA, so that's where my cash in a brokerage account is going to be parked. If I were not comfortable with it, I'd probably use the SEC yield of Treasury MM.
I understand that. My question was not about what funds you used but why you used the MM funds as zero duration instruments. I don't buy MM instruments so I don't pay a lot of attention to them but you made me look. CA Muni has an average maturity of 29 days. VG Federal and Prime MM funds have a duration of 50 days. If you use the returns (SEC yield) from these MM instruments but used the actual maturity instead of zero duration would you still have the negative slope at the short end of your curve?
Perhaps not. The duration of the holdings is greater than 0, but since the funds maintain a NAV of $1/share, they are 0-year from my perspective, since there is no term risk to me.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Tue Apr 24, 2018 11:18 pm

One guideline suggested by Larry Swedroe is to only extend maturity if you earn at least an additional 20 basis points (bps) of yield per extra year of maturity. Of course the more the better, since a steeper yield curve provides a bigger buffer from capital loss due to yields at shorter maturities increasing. For example, with a 2-year yield at 2.5% and a 3-year yield at 2.7% (20 bps higher), the 2-year yield can increase by 20 bps in one year before you experience a negative 1-year capital return for the 3-year security, assuming bonds or CDs are bought at par (100, or 100% of face value).

So how do the yield curves look relative to this guideline? Yields have changed a bit since I posted the chart in the OP, and you can't really read these numbers off the chart, but the following observations are based on the latest quotes from Fidelity.

Using the raw Treasury yield, so applicable in an IRA and assuming I can actually get the quoted yield, at the 1-year yield of 2.28%, I'd earn 48 bps more than the Prime MM 0-year yield I use of 1.80%. This is way more than 20 bps, so it appears that a 1-year Treasury makes sense in an IRA when compared to Prime MM. With the 2-year yield at 2.50%, the additional yield is only 22 bps, so this just barely makes the cut. The 3-year yield of 2.66% only provides an extra 16 bps, so extending maturity from 2-year to 3-year does not pass the test. Things get even bleaker at longer maturities.

The CD yield curve is not quite as steep at the short end, providing 35 bps for extending from 0-year Prime MM at 1.80% to 1-year CD at 2.15%; still good, but obviously the 1-year Treasury is better in an IRA. But the 2-year CD at 2.75% provides 60 bps of more yield than the 1-year CD, so this appears to be a much better deal. Or you could say it provides 95 bps over the 0-year, or 47.5 bps per year, so still a good deal from this perspective. The 3-year CD at 2.90% provides only 15 bps more yield than the 2-year, so it doesn't make the cut. If you can find a 3-year on the secondary market at about 3%, then that's 25 bps more than the 2-year, so unless you can find higher 2-year yield in the secondary market, extending to 3-year maturity would make the cut.

Extending CD maturity to 5-year at 3.15% only provides 25 bps more than 3-year at 2.90%, so only 12.5 bps/year--not nearly good enough for me.

This reply is focused mostly on how much extra yield you get by extending maturity for a particular type of fixed-income security, but I obviously also want to compare between the different types of securities (unless my investment policy restricts me from doing so--e.g., if I'm only going to buy US Treasuries no matter what CDs offer in an IRA, or AA or AAA munis offer in taxable). So even though the CD yield curve might be steeper from 1-year to 2-year maturity than the Treasury TEY yield curve in taxable, I'll obviously prefer the 2-year Treasury that has higher TEY than the 2-year CD (but I'll also look at AA munis at 2-year maturity in taxable).

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Doc
Posts: 8614
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Doc » Wed Apr 25, 2018 10:04 am

Kevin M wrote:
Tue Apr 24, 2018 11:18 pm
One guideline suggested by Larry Swedroe is to only extend maturity if you earn at least an additional 20 basis points (bps) of yield per extra year of maturity.
I finally took the time to search for Larry's recommendation.
First, for TAXABLE bonds I have suggested 20bp per year as a good starting point for extending maturity. For munis then it would be more like 15-16
So I found out where I got the 15 bps from. It's Munis
Second, DFA uses a shifting maturity approach to managing bonds which is based on original work by Fama. It's based on best predictor of future curves is today's.
Third, I would add this. Having a rule is what is important. Not the particular number you use--like with rebalancing. The more risk of unexpected inflation, the higher the bar should be, and vice versa.
viewtopic.php?p=1826872&sid=1b53a451694 ... 2#p1826872

Hey Kevin, where were you in October 2013? Did you take a vacation from the Bogleheads? There are no posts from you in that thread. :D
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

User avatar
Doc
Posts: 8614
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Doc » Wed Apr 25, 2018 11:44 am

Kevin M wrote:
Tue Apr 24, 2018 11:18 pm
So how do the yield curves look relative to this guideline? Yields have changed a bit since I posted the chart in the OP, and you can't really read these numbers off the chart ...
For Treasuries here's a chart showing the change in the curve at several times over the last few years:

Image

The data constant maturity rates from Treasury.gov https://www.treasury.gov/resource-cente ... data=yield

(I don't like the Treasuries yield chart presentation because of the nonlinear time scale they use.)
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

User avatar
randomizer
Posts: 1521
Joined: Sun Jul 06, 2014 3:46 pm

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by randomizer » Wed Apr 25, 2018 11:52 am

Very useful and informative.
87.5:12.5, EM tilt — HODL the course!

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Wed Apr 25, 2018 1:31 pm

Doc wrote:
Wed Apr 25, 2018 10:04 am
Kevin M wrote:
Tue Apr 24, 2018 11:18 pm
One guideline suggested by Larry Swedroe is to only extend maturity if you earn at least an additional 20 basis points (bps) of yield per extra year of maturity.
I finally took the time to search for Larry's recommendation.
First, for TAXABLE bonds I have suggested 20bp per year as a good starting point for extending maturity. For munis then it would be more like 15-16
So I found out where I got the 15 bps from. It's Munis
Yep, I thought I mentioned that in one of our other discussions in another thread. Larry repeated the 15 bps/year guideline for munis in an email in which he answered some questions I had about munis as I was starting to buy them late last year.

However, what I do is convert to TEY, then look for at least 20 bps/year of TEY.
Second, DFA uses a shifting maturity approach to managing bonds which is based on original work by Fama. It's based on best predictor of future curves is today's.
Yeah, you're basically looking for the steepest segment of the yield curve consistent your investment policy on maturity range. A fund has to buy in less steep segments if their policy is to hold maturities in those ranges, but I don't have to do that, which is why most bond funds don't look attractive to me with the yield curve very steep at the short end and pretty flat beyond that).
Third, I would add this. Having a rule is what is important. Not the particular number you use--like with rebalancing. The more risk of unexpected inflation, the higher the bar should be, and vice versa.
Definitely! With a 2-year CD maturing 5/4/2020 providing almost 47 bps/year over the 0-year yield of 1.80% (that I use), I'm more picky now.

For example, today I bought a CD maturing 02/10/2021 with net yield of 2.994%, which comes to about 43 bps/year relative to 0-year at 1.80%, and about 32 bps/year relative to 2-year at 2.75%. This also is more than 9 bps over the 3-year at 2.90% maturing 5/4/2021, so a decent yield premium at shorter maturity.

Unfortunately, there was only 1 available, so I bought it in a small Roth IRA where I only had enough cash left to buy one anyway (until proceeds from my maturing PenFed Roth IRA CD arrive in early May--assuming no kinks with the transfer). This is an example of the small-lot anomaly that Larry has mentioned, and I have found to be the case both for CDs and munis (of course not for Treasuries, where it works the opposite way).
viewtopic.php?p=1826872&sid=1b53a451694 ... 2#p1826872

Hey Kevin, where were you in October 2013? Did you take a vacation from the Bogleheads? There are no posts from you in that thread. :D
Back then I was buying direct CDs, mostly 5-year, and earning yield premiums of 100 bps or more over Treasuries of same maturity with early withdrawal penalties of six months of interest, so I wasn't paying much attention to the maturity extension guidelines. Since the small early penalty limits your term risk to much less than a 5-year brokered fixed-income security, the guidelines didn't really apply to me then.

But thanks for the link. I'll scan through that thread.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

international001
Posts: 440
Joined: Thu Feb 15, 2018 7:31 pm

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by international001 » Thu Apr 26, 2018 5:13 am

Newbie question: Why are CD yields higher than treasuries? Are they considered less safe? Are we assuming those are not FDIC protected?

User avatar
Doc
Posts: 8614
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Doc » Thu Apr 26, 2018 9:50 am

international001 wrote:
Thu Apr 26, 2018 5:13 am
Newbie question: Why are CD yields higher than treasuries? Are they considered less safe? Are we assuming those are not FDIC protected?
CD's are not as liquid as Treasuries. So if you want or need to sell before maturity it is more difficult and often more expensive to sell the CD. Investors demand a price for this lower liquidity i.e. higher yield.

Kevin's analysis is based on holding to maturity. I hardly ever hold Treasuries to maturity which is why Kevin and I often come up with different conclusions to what appears to be the same situation.
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Thu Apr 26, 2018 12:19 pm

Doc wrote:
Thu Apr 26, 2018 9:50 am
international001 wrote:
Thu Apr 26, 2018 5:13 am
Newbie question: Why are CD yields higher than treasuries? Are they considered less safe? Are we assuming those are not FDIC protected?
CD's are not as liquid as Treasuries. So if you want or need to sell before maturity it is more difficult and often more expensive to sell the CD. Investors demand a price for this lower liquidity i.e. higher yield.

Kevin's analysis is based on holding to maturity. I hardly ever hold Treasuries to maturity which is why Kevin and I often come up with different conclusions to what appears to be the same situation.
Yes, liquidity is one reason. Of 226 2-3 year CDs I evaluated today, 200 had bid quotes. The average spread was 0.84% (ask/bid - 1), the minimum spread was 0.05%, and maximum was 1.54%. Spreads generally increase with increasing maturity, but it is very erratic. Compare this to spreads of a few basis points for Treasuries, so as Doc says, if you don't plan to hold to maturity, you need to factor in the bid/ask spread, and this is likely to make Treasuries a better choice.

Another reason is that institutional investors can't take advantage of FDIC insurance, so CDs aren't essentially risk free to them as they are to smaller investors.

Related to liquidity and institutional buying is the sheer size of the Treasury market compared to the CD market--at least based on what I see at Fidelity. Largest quantity offered of any one CD is 250 (probably because of the $250K FDIC insurance limit), while Treasuries are offered in quantities as large as 25,000. So there's much more money competing for Treasuries than for CDs, which I'd think would drive prices up relative to CDs.

Another advantage of Treasuries is the state tax deduction, so this would tend to drive yields down some, since TEYs of Treasuries can be competitive with CDs if you pay state income tax. Treasury TEYs are slightly higher than new-issue CD yields out to 5-year maturity for me.

Still, note that Treasury yields are higher than new-issue CD yields out to 1-year maturity, even without considering state tax exemption, so even in a tax-advantaged account. So it's not true that CD yields are higher than Treasury yields at all maturities, and the yield premiums of CDs over Treasuries at longer than 1-year maturity have shrunk a lot over the last year or so.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Thu Apr 26, 2018 2:05 pm

The yield curves in the OP just reflect a sampling of the highest yields at the maturities that Fidelity shows on its yield summary page. The yields may not represent what you actually can get at smaller quantities for Treasuries, may be for maturities as much as one month longer than indicated, and don't reflect net yields after commissions for CDs and munis. There also is no indication of taxable-equivalent yield (TEY), which for me would be different for a CA muni with the same net yield as a non-CA muni.

The first set of charts below show net TEY for me for CA and non-CA munis. I've placed them side by side so it's easier to compare TEY for CA and non-CA munis, helping me to zero in on which look more attractive at various maturities.

In my muni search criteria, I search for 3-month to 39-month munis, but set a minimum yield that will exclude most munis at shorter maturities. Commissions cut into shorter-term securities, which is why the net TEY may be below what I actually would accept, but at somewhat longer maturity, the same before-commission yield could produce an acceptable net TEY. Munis in these results must be at least Moody's Aa2 or S&P AA. In the non-CA munis, I've also excluded IL because these tend to have higher yields (reflecting higher risk), and I already have as much in IL munis as I want at this point.

The red curve in each chart is a trend line.

Image

There obviously are a lot less CA munis, and note that the spread between low yield and high yield at a given maturity is much less than for non-CA munis. In general I'll prefer a CA muni, since the state income tax reporting is easier, but I also don't want too much concentration in CA.

CA munis tend to be more attractive at the shorter terms. Note that non-CA munis start to exceed 3% net TEY at about 1.7-year maturity, while CA munis barely touch 3% at 2.5-year maturity. Of course this is only an initial screen, and the next step is to examine more closely the munis that have the higher bps/year for extending maturity.

For example, for the 1.7-year muni that exceeds 3%, there are only 5 available. I might buy 5, but typically I'd prefer minimum quantity of 10. Other than that, it looks OK. It's an IN revenue bond rated AA by S&P. It provides about 55 bps/year of yield relative to the shortest-term muni in the search results (0.3-year at 2.28%), and about 46 bps/year relative to the CA muni MM fund at 2.26% TEY.

Below is a similar chart for CDs that have yield of at least 2.8%. This is to find secondary CDs that will provide at least the 2.75% net yield of a new-issue 2-year CD. I don't bother looking for anything less, since I'll probably beat the CD with a Treasury at the shorter maturities. But I guess I should at least look to make sure. Maybe I'll start doing that.

Image

As in the chart in the OP, the yield curve is quite steep out to 2-year maturity, but flattens out significantly beyond that.

For those who want to go out to about 3-year maturity, I see CUSIP 949763FW1 from Wells Fargo, at 2.97% net for 2.93-year maturity. This is about 24 bps/year relative to 2-year new-issue at 2.75%, about 40 bps/year relative to 0-year Prime MM at 1.81%; it's also 7 bps more than 3-year new issue at 2.90% and a bit shorter maturity. There were 886 available with minimum quantity 1 when I pulled the data earlier today, so looks pretty good for about 3-year maturity.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Fri Apr 27, 2018 2:25 pm

Kevin M wrote:
Thu Apr 26, 2018 2:05 pm
Below is a similar chart for CDs that have yield of at least 2.8%. This is to find secondary CDs that will provide at least the 2.75% net yield of a new-issue 2-year CD. I don't bother looking for anything less, since I'll probably beat the CD with a Treasury at the shorter maturities. But I guess I should at least look to make sure. Maybe I'll start doing that.

Image
Today I lowered my minimum gross yield (ask yield before commission) to 1.9%, and generated the following CD yield curve from the results.

Image

This shows CDs with maturities shorter than two years, and also shows much more divergence of yields in the 2-year to 3+-year range--you are rewarded for selectivity.

Note that even with a gross yield of 1.9% or more, the commission cuts into the yield too much for these shorter-maturities to even beat Prime MM at 1.81%. For example, a 0.43-year CD at 1.93% is only 1.69% after $1/CD commission.

To get at least 20 bps/year of yield relative to Prime MM at 1.81%, you have to go out a bit beyond 9-month maturity. There is a 0.78-year CD with net yield of about 2%, which comes to about 24 bps/year of extra yield relative to Prime MM. Maturity of this CD is 2/5/2019. Closest maturity Treasury is 1/31/2019, and for quantity 10 yield is 2.13%, so as expected, Treasury beats CD at this shorter maturity (since there is no commission for Treasury, we can just go by quoted ask yield).

You probably are better off with new-issue CDs at these shorter maturities, but as with Treasuries, the yields shown on the summary page may be for maturities onger than indicated. The yield shown for 6-month CD is 1.95%, but this is for a CD that matures 11/13/2018 and settles 5/11/2018, so it's really a 0.55-year CD counting from today, and you can't buy it for two weeks. A CD that matures 11/2/2018 and settles 5/2/2018 yields 1.85%. I see a Treasury maturing 10/31/2018 that yields 1.93% for minimum quantity 1 and 1.94% for minimum quantity 25. So again, Treasuries are better at these shorter maturities.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
ps56k
Posts: 804
Joined: Sat Mar 19, 2011 1:28 pm
Location: Chicago area

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by ps56k » Fri Apr 27, 2018 3:18 pm

Interesting following along..... learning.... scrolling up/down -
Wonder how this very specific targeting would compare against the backdrop of the usual Vanguard Total Bond Index fund ?

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Fri Apr 27, 2018 4:05 pm

ps56k wrote:
Fri Apr 27, 2018 3:18 pm
Interesting following along..... learning.... scrolling up/down -
Wonder how this very specific targeting would compare against the backdrop of the usual Vanguard Total Bond Index fund ?
I don't like total bond for a number of reasons, for example, it holds lots of Treasuries, and I can beat Treasuries with CDs at maturities beyond one year, assuming I hold to maturity (which I do). If I want something like total bond, I would hold CDs instead of the Treasuries, and then supplement with an investment-grade or corporate bond fund to get similar credit risk and some more term risk (and I do hold such funds in IRAs, along with some CA muni funds in taxable).

I don't like bond funds in general now because they generally hold bonds beyond the steep ranges of the yield curves--even something like limited-term tax-exempt holds about 23% of bonds with maturity of five years or more and about 40% with maturity greater than three years. Of course intermediate-term funds hold even more at longer maturities where yield curves are quite flat. Little expected reward for taking on the additional term risk.

Having said that, I continue to hold the bond funds I already had--I'm just not adding to them. I've had about 25-30% of fixed income in intermediate-term investment-grade or muni bond funds for some years, and probably will continue with that to hedge the possibility of intermediate-term yields staying low for some time, and for the possible reward of taking some credit risk.

For similar reasons, I did add a relatively small holding of limited-term tax exempt in January. If I'm going to extend maturity beyond three years and roll the bonds, might as well use a fund, although there still are advantages to holding individual munis, but I have another thread going on that.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

Mitchell777
Posts: 1022
Joined: Mon May 21, 2007 6:32 am

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Mitchell777 » Fri Apr 27, 2018 5:32 pm

Sorry if I missed it but what is the cost to sell a Treasury? I see Vanguard does not handle the sale but will provide access to an OTC market. I know the bid/ask part just not sure what additional costs there may be. I'm not sure I want to sell but I bought a 5 year note a month or so ago at 2.625% (low state tax rate) and I'm leaning right now toward the 2 and 3 year CD's. Thanks

stlutz
Posts: 4742
Joined: Fri Jan 02, 2009 1:08 am

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by stlutz » Fri Apr 27, 2018 6:32 pm

Sorry if I missed it but what is the cost to sell a Treasury?
At Vanguard, E*Trade, Schwab, and Fidelity the cost is $0.

stlutz
Posts: 4742
Joined: Fri Jan 02, 2009 1:08 am

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by stlutz » Fri Apr 27, 2018 6:35 pm

I don't like bond funds in general now because they generally hold bonds beyond the steep ranges of the yield curves--even something like limited-term tax-exempt holds about 23% of bonds with maturity of five years or more and about 40% with maturity greater than three years. Of course intermediate-term funds hold even more at longer maturities where yield curves are quite flat. Little expected reward for taking on the additional term risk.
What do you think the odds of a recession in the next 5 years are? If we have one, which direction do you think rates will go? Assuming the answer is "down", wouldn't it have been a good idea to lock in the higher rates now? Sticking with short-term bonds has the risk that you might not be able to get an equivalent yield to today in 2 years when your bond matures.

That's what the bond market is telling you now with the current shape of the yield curve. It might be wrong, but it's always good to start out with an understanding of what the market is saying first.

gips
Posts: 475
Joined: Mon May 13, 2013 5:42 pm

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by gips » Fri Apr 27, 2018 6:52 pm

Kevin M wrote:
Tue Apr 24, 2018 11:18 pm
One guideline suggested by Larry Swedroe is to only extend maturity if you earn at least an additional 20 basis points (bps) of yield per extra year of maturity.
Kevin, thank you for the good analysis. I've started thinking about larry's 20bps rule. Do you think Larry's rule makes sense in every market? with every shape of the yield of the curve? Mostly, I'm wondering if the rule makes sense in our current environment where we can be pretty sure the fed will increase rates at least two more times this year.

It's too bad larry's doesn't post here anymore...

gips
Posts: 475
Joined: Mon May 13, 2013 5:42 pm

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by gips » Fri Apr 27, 2018 6:52 pm

Kevin M wrote:
Tue Apr 24, 2018 11:18 pm
One guideline suggested by Larry Swedroe is to only extend maturity if you earn at least an additional 20 basis points (bps) of yield per extra year of maturity.
Kevin, thank you for the good analysis. I've started thinking about larry's 20bps rule. Do you think Larry's rule makes sense in every market? with every shape of the yield of the curve? Mostly, I'm wondering if the rule makes sense in our current environment where we can be pretty sure the fed will increase rates at least two more times this year.

It's too bad larry' doesn't post here anymore or we could just ask him!

stlutz
Posts: 4742
Joined: Fri Jan 02, 2009 1:08 am

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by stlutz » Fri Apr 27, 2018 7:02 pm

I've started thinking about larry's 20bps rule. Do you think Larry's rule makes sense in every market? with every shape of the yield of the curve? Mostly, I'm wondering if the rule makes sense in our current environment where we can be pretty sure the fed will increase rates at least two more times this year.
One thing to keep in mind with this "rule" is that it's based on the average maturity of the portfolio. So, if the rule is currently saying that your average maturity should be 2.5 years, that would mean that the new bonds you are buying probably have a maturity closer to 5 years if you are rolling a bond ladder.

I've personally backtested various approaches to variable maturity rules and my conclusion is that they work extremely well--up until about 20 years ago beyond which there really hasn't been a benefit vs. using a fixed average across time. I haven't published any peer-reviewed papers, so take my comments for what they are worth, but one can use longinvests bond calculator spreadsheet to do their own modeling.

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Fri Apr 27, 2018 8:47 pm

stlutz wrote:
Fri Apr 27, 2018 6:35 pm
What do you think the odds of a recession in the next 5 years are? If we have one, which direction do you think rates will go?
I really don't know. But if the yields fall during a recession that starts a few years from now, they might fall from levels that are significantly higher than they are now, and they might end up the same as or even higher than they are now. If we look back to a period when yields were generally rising, that's the way it looks to me if we look at the 10-year Treasury, for example:

Image
Assuming the answer is "down", wouldn't it have been a good idea to lock in the higher rates now?
But as shown in the chart above, the rates now might not be higher than at the end of the next recession, so we don't really know that we're locking in "high rates" now.
Sticking with short-term bonds has the risk that you might not be able to get an equivalent yield to today in 2 years when your bond matures.
True. With shorter-term bonds you have more reinvestment risk, and with longer-term bonds you have more term risk. I prefer the reinvestment risk to the term risk, given how flat the yield curves are beyond 2 or 3 years. The 25-30% of my fixed-income in intermediate-term bond funds (but not Treasury or Treasury-heavy funds) provides some hedge against lower intermediate-term yields in 2-3 years, so it's not all or nothing.
That's what the bond market is telling you now with the current shape of the yield curve. It might be wrong, but it's always good to start out with an understanding of what the market is saying first.
I think you're referring to the expectations hypothesis of the term structure of interest rates. If this were the only explanation of the yield curve, then we should expect to do about the same rolling shorter-term bonds as holding longer-term bonds, and the yield curve would never be telling us which is better. One problem is that the empirical evidence does not support the expectations hypothesis, or so I've read.

Another explanation for the shape of the yield curve is liquidity preference theory, which basically says there is a risk premium for taking term risk, since there is more uncertainty in expected return at longer maturities. The less steep the yield curve, the less risk premium for taking the term risk. I even saw post a few months ago referencing an academic perspective that indicated that the term premium was negative--can't remember the details, but I think it might have been lack_ey that posted it.

I'm getting about the same yield in a 2-year CD as in a 5-year Treasury in a tax-advantaged account. If I wanted to "lock in" a higher yield, I'd use a 5-year CD with a yield premium of about 35 basis points over 5-year Treasury (3.15% - 2.8%), or a 10-year CD with a yield premium of about 50 bps over the 10-year Treasury (3.45% - 2.95%)--again, in a tax-advantaged account, and if planning to hold to maturity (and either match liabilities, roll at maturity, or some combination thereof).

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Fri Apr 27, 2018 9:01 pm

gips wrote:
Fri Apr 27, 2018 6:52 pm
Kevin M wrote:
Tue Apr 24, 2018 11:18 pm
One guideline suggested by Larry Swedroe is to only extend maturity if you earn at least an additional 20 basis points (bps) of yield per extra year of maturity.
Kevin, thank you for the good analysis. I've started thinking about larry's 20bps rule. Do you think Larry's rule makes sense in every market? with every shape of the yield of the curve? Mostly, I'm wondering if the rule makes sense in our current environment where we can be pretty sure the fed will increase rates at least two more times this year.

It's too bad larry's doesn't post here anymore...
First, it's not a rule, it's a guideline. Second, Larry says that it's better to have some guideline or policy than none, regardless of the actual values you use.

With the 2-year CD at 2.75% providing 60 bps/year more than the 1-year CD at 2.15%, or about 47 bps/year relative to 0-year Prime MM at 1.81%, I'm hesitant to extend maturity beyond two years unless I can get 25-30 bps/year relative to the 2-year and say 35-40 bps/year relative to the 0-year yield. So I'd say I buy at most five 3-year CDs for every 10 2-year CDs lately.

The fed controls the federal funds rate (FFR), and the FFR doesn't have much bearing on Treasury yields for maturities of more than about 1-month. In other words, even if the FFR is increased per current guidance, that doesn't mean that 2-year or 5-year or 10-year yield also will increase by a similar amount, or even at all. And who knows if the FFR will actually be increased as currently projected. In other words, as Larry says, there are no good forecasters, and that applies to fixed-income yields as much as anything else.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Fri Apr 27, 2018 9:15 pm

stlutz wrote:
Fri Apr 27, 2018 7:02 pm
One thing to keep in mind with this "rule" is that it's based on the average maturity of the portfolio. So, if the rule is currently saying that your average maturity should be 2.5 years, that would mean that the new bonds you are buying probably have a maturity closer to 5 years if you are rolling a bond ladder.
That's not the way I look at it.

If you buy a 5-year bond at par, and the 4-year bond currently yields 20 bps less, then the 4-year yield can increase 20 bps before your 1-year capital return component goes negative. So you want some positive slope to provide a positive expected capital return component if you want to assume a static yield curve to estimate your expected return, but realizing that the yield curve is unlikely to remain static, which is why there is uncertainty in the capital return component. The steeper the yield curve the more the 4-year yield has to increase before your capital return component goes negative.

Of course yields generally have risen by much more than 20 bps in the last year, which is why the capital return component of a bond fund or bond ladder probably is negative, and why we see such low (and even negative) 1-year bond fund total returns. Heck, most yields of 2-year maturity or more have risen by more than 20 bps just during this month.

Buying individual fixed-income securities and keeping maturities short doesn't save you from some of this pain. We have all lost value in fixed-income at maturities longer than one month lately. But we can roll the shorter maturities over sooner, which of course is the benefit of keeping maturities shorter if yields continue to rise (but I'm not predicting that they will).

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
patrick013
Posts: 2367
Joined: Mon Jul 13, 2015 7:49 pm

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by patrick013 » Fri Apr 27, 2018 9:30 pm

Current Interest Rate information

Realistically if FFR rates rise to +3% I think 5 year TRSY's will
have the healthiest spreads compared to their mean average. There
is still some lingering doubt about 10 year and longer spreads.
Will 2025 bring a FFR rate decrease making LT bond pricing too
low in hindsight ? So pricing may remain high and spreads low
thru that evaluation period. I don't think the FED wants a low
interest rate economy but any reason to keep the prices on the LT
bonds high with spreads at lower than their mean average I think
the market will do that.

Rates are presently higher and my target date to go longer term is
per the graph above. I'm expecting a .88 to .93 correlation of the
FFR to the TRSY yield movement going forward per historical data.

Have a good one.
age in bonds, buy-and-hold, 10 year business cycle

protagonist
Posts: 5388
Joined: Sun Dec 26, 2010 12:47 pm

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by protagonist » Fri Apr 27, 2018 9:36 pm

Hi Kevin.

First, thanks for a great post and thoughful research.

In your humble opinion, with bank accounts offering 2% interest, at what point do you think investing in a CD or Treasury is worth it today?

Is it worth the lack of liquidity to milk an extra 1/2-1% , especially with rising rates? It certainly seems like rates will be higher in another year or so.

I hope to come into a small (taxable) windfall in about a month (selling a condo), so this is relevant to me.

Thanks.
Last edited by protagonist on Fri Apr 27, 2018 9:41 pm, edited 1 time in total.

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Fri Apr 27, 2018 9:45 pm

protagonist wrote:
Fri Apr 27, 2018 9:36 pm
Hi Kevin.

First, thanks for a great post and thoughful research.

In your humble opinion, with bank accounts offering 2% interest, at what point do you think investing in a CD or Treasury is worth it today?
In taxable, my TEY for CA muni MM fund is 2.25%, so that's my 0-year hurdle in taxable. I can't use a 2% savings account in an IRA--at least not conveniently, so the hurdle there is Prime MM at 1.81%.

In taxable, with a 1-year AA muni at about 2.8% TEY, that gives me about 55 bps/year of extra yield, which is quite steep, so that seems decent compensation for taking a little term risk. One-year Treasury TEY for me is about 2.5%, so the AA muni looks better, given historical 1-year default rates of 0%.
Is it worth the lack of liquidity to milk an extra 1/2-1% , especially with rising rates? It certainly seems like rates will be higher in another year or so.
They certainly have been rising, but we don't know that they will continue to rise, or if they do, by how much. It has been paying to be patient, and gradually deploy cash into bonds or CDs. I have quite a bit more CDs maturing soon, so will continue to deploy into CDs, Treasuries, and munis, depending on whether it's taxable or tax-advantaged. I've been overweighting 2-year maturities, but will see how things look when the cash becomes available.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

perl
Posts: 130
Joined: Mon Apr 28, 2014 12:46 pm

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by perl » Sat Apr 28, 2018 4:37 am

Can you say something about how you choose and buy brokered CDs? Do you use Vanguard or another platform? Once you've identified a term you want, do you just pick the highest rate with a reasonable spread? Or does the issuing bank matter?

lazyday
Posts: 3297
Joined: Wed Mar 14, 2007 10:27 pm

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by lazyday » Sat Apr 28, 2018 5:01 am

Kevin M wrote:
Fri Apr 27, 2018 8:47 pm
I even saw post a few months ago referencing an academic perspective that indicated that the term premium was negative--can't remember the details, but I think it might have been lack_ey that posted it.
Here it is: viewtopic.php?t=230547

As of the end of March, it might be the case that short duration Treasuries have higher expected return than longer duration, given expectations for rate rises.

Recent data xls: https://www.newyorkfed.org/medialibrary ... remium.xls

Explanation for column M header: viewtopic.php?t=230547#p3624202

txranger
Posts: 100
Joined: Sun Jun 11, 2017 9:14 am

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by txranger » Sat Apr 28, 2018 6:48 am

How quickly can you get cash out of vanguard prime and muni mm? My spending accounts r at Fido and Schwab. Schwab is 1 biz day delay to get out of muni mm and it trails vanguard by 10 bp. Fido is immediate liquity which is great — can pull cash out of atm but it’s muni funds yield 1.3%, a lot lower.

For vanguard looks like u can only draw immediate ach on federal mm?

I pulled cash out of ally and alliant as yield is not as good.
Tnx appr

User avatar
Doc
Posts: 8614
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Doc » Sat Apr 28, 2018 11:07 am

I updated my yield chart for more recent data.

Image
Kevin M wrote:
Fri Apr 27, 2018 9:01 pm
The fed controls the federal funds rate (FFR), and the FFR doesn't have much bearing on Treasury yields for maturities of more than about 1-month. In other words, even if the FFR is increased per current guidance, that doesn't mean that 2-year or 5-year or 10-year yield also will increase by a similar amount, or even at all. And who knows if the FFR will actually be increased as currently projected. In other words, as Larry says, there are no good forecasters, and that applies to fixed-income yields as much as anything else.
Agreed that the FFR doesn't have much bearing on longer maturities but that doesn't mean that Fed actions do not influence longer rates. The Fed has a gazillion in Treasuries left on its balance sheet from the prior stimulus. The Fed is now in the process of quantitative easing - getting rid of that balance sheet excess. How fast it does that does influence longer term rates. If the Fed holds $100 million in five year notes maturing in June does it buy another $100 million to replace them or only $50 million or none at all. That decision is going to effect the price of the new five.

Looking at the chart above you can get some idea of what is happening.

The difference between the five year and the one month yields are as follows:

April '18: 1.19
January '18: 1.17 (little change)
November '17: 0.98 (shape changer)
September '17: 0.87 (some shape change)

Late last year the shape of the curve was changing. This year it seem to be moving up the same amount across the maturity spectrum.

In my opinion all this makes questionable Kevin's assumption of holding to maturity as being the default case. Holding to maturity negates the benefits for "riding the yield" curve no matter how difficult it is hard it is to predict the actual value of the effect.
DEFINITION of 'Riding the Yield Curve'
Riding the Yield Curve is a trading strategy that involves buying a long-term bond and selling it before it matures so as to profit from the declining yield that occurs over the life of a bond. Investors hope to achieve capital gains by employing this strategy.
Read more: Riding the Yield Curve https://www.investopedia.com/terms/r/ri ... z5Dz45uxTs
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Sat Apr 28, 2018 1:10 pm

The chart I posted earlier with minimum gross yield (ask yield before commission) of at least 1.9% had too many CDs to effectively evaluate (1,353), and included mostly CDs with non-competitive yields. Here it is again:

Image

So I added a filter in my spreadsheet that only included a CD if its yield was higher than the previous 10 CDs (sorted by maturity), with the intention of slicing off the peaks of the noisy chart above. Here's the result:

Image

This reduces the number of CDs to 122, and makes the chart much easier to scan and pick out CDs that have relatively attractive yields at a given maturity.

Looking at this new chart, it looks like I could filter even more aggressively, for example yield must by higher than previous 20 CDs, further reducing the number of CDs to evaluate and still not missing the best ones. I just tried that, and it looks even better (smoother), and still catches the best deals.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Sat Apr 28, 2018 1:56 pm

perl wrote:
Sat Apr 28, 2018 4:37 am
Can you say something about how you choose and buy brokered CDs? Do you use Vanguard or another platform? Once you've identified a term you want, do you just pick the highest rate with a reasonable spread? Or does the issuing bank matter?
I discuss it at length in this thread: 2-year CDs at Vanguard and Fidelity Today. In that post, I started with the intention of just comparing secondary market brokered 2-year CDs at Vanguard and Fidelity on that particular day, but the thread evolved into a broader discussion.

I'll summarize my process here.

I use both Vanguard and Fidelity, but have been favoring Fidelity lately. The Fidelity feature to download search results into a spreadsheet makes finding the good CD deals much more effective (and enabled me to produce the charts in this thread). I haven't seen that functionality at Vanguard. Sometimes when I find a good CD at Fidelity, but want to buy it in an account at Vanguard, I'll just jump to Vanguard and search for the CUSIP. Sometimes Vanguard offers the same CD at same price, and I'll jump on it (if it's not gone by then), but sometimes they don't offer it at all, or they offer it at a higher price.

Not sure what you mean by "reasonable spread", but if you're talking about the bid/ask spread, I ignore it. I only care about ask price/yield, since that's the offer on the table, and what matters is how it compares to alternatives I can buy. I wouldn't buy a brokered CD unless planning to hold to maturity, since bid/ask spreads generally are quite high (about 0.8% average for 2-3 year CDs last time I checked).

I first check new-issue yields for the maturities of interest, and then I look for secondary market CDs that will beat those yields. Or in a small account (I am an agent for some smaller family member accounts), I might buy it even if it meets the new issue yield, since new-issue minimum quantity is 10 ($10,000 face), and in a smaller account there might not be enough to buy 10.

I also check to make sure the net yield beats the Treasury yield of same maturity in an IRA, which it generally doesn't at maturities of less than two years currently. If in taxable, I compare to taxable-equivalent yield (TEY) of Treasuries and AA munis of same maturities. Generally Treasuries beat CDs and AA munis beat Treasuries in taxable at my marginal tax rates.

Walking through an IRA example using latest chart posted above, I see a 1.99-year CD at net yield of 2.754% (can't read this precision off the chart, but I can in my spreadsheet). From the chart it's clear that this is the best deal available. It beats the new-issue 2-year yield of 2.75%--just barely, but it's a bit shorter maturity. I already know that 2-year CDs beat 2-year Treasuries in IRA by about 25 basis points, so no need to look at Treasuries in detail. However, looking at the details for this CD, there were only 2 available ($2,000 face value), so good for a small account, but doesn't do me much good if I'm looking for minimum quantity of 10. This is an example of what I'd buy in a small account (and did with this one or something similar last week).

Nothing else is competitive with new-issue 2-year, and the earliest settlement for that at 2.75% is next Thursday, so it probably still will be available next Tuesday. In this case, if I were buying on Monday, I'd probably wait until Tuesday to pick up the new issue if I didn't find anything better on secondary market by then. If I'm getting toward the end of deploying whatever chunk of cash I've been working on, I'll be even more selective, and look for at least 5 basis point yield premium over new-issue, as I've typically been able to do so. Often the new issue trades at a higher net yield just after and even just before or on the settlement date.

All CDs offered by Fidelity and Vanguard are FDIC insured, so issuing bank only matters if approaching $250K in one bank in an IRA, or $250K per beneficiary in taxable, since my taxable accounts are trust accounts. This is across all accounts, for brokered CDs and CDs held directly at the bank. At some point I'll have to start paying attention to this in IRAs, but I'm not there yet; I still have lots in direct IRA CDs, and much of it at credit unions, so no overlap with brokered bank CDs, but as they mature, I transfer to Fidelity (previously Vanguard)--as long as the brokered CDs still look most attractive to me.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Sat Apr 28, 2018 2:08 pm

txranger wrote:
Sat Apr 28, 2018 6:48 am
How quickly can you get cash out of vanguard prime and muni mm? My spending accounts r at Fido and Schwab. Schwab is 1 biz day delay to get out of muni mm and it trails vanguard by 10 bp. Fido is immediate liquity which is great — can pull cash out of atm but it’s muni funds yield 1.3%, a lot lower.

For vanguard looks like u can only draw immediate ach on federal mm?

I pulled cash out of ally and alliant as yield is not as good.
Tnx appr
Good question. Of course this is all relevant only to taxable, since IRA transfers take too long to make it feasible to move cash back and forth between IRA accounts. So for me, this is more relevant to buying munis, since that's mostly what I've been buying in taxable at Fidelity.

Since settlement for CDs and munis is T+2, you can place your order and you're good as long as you get cash there by second business day after trade. I already had my Ally bank account linked to Fidelity, so while cash was at Ally, I would place the trade, then place a bank transfer order to get cash from Ally to Fidelity settlement fund, and it would always arrive in time.

When I moved cash to VG CA muni MM for much higher TEY than Ally savings, I had to set something else up. So I opened a cash management account (CMA) at Fidelity, and then set up the link from Vanguard to Fidelity. Now I can place trade at Fidelity, then at Vanguard sell the exact dollar amount to cover the trade, setting Fidelity CMA as the receiving bank. Cash arrives there in T+2 at most--maybe even T+1, and then I transfer the cash from CMA to brokerage settlement fund.

Only problem with this was that Vanguard places a 7-day hold on new cash in the MM fund, and a 7-10 day hold on the ability to transfer to a newly linked bank (in this case, the Fidelity CMA). I was able to get around this by spending some time with Vanguard on the phone, as I wanted to buy some munis right away. The Vanguard rep called Ally to verify that the funds would clear, then they asked me some security questions to verify the new bank (this was a one-time deal--I still had to wait the 7-10 days for the CMA to be verified for subsequent transfers).

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Doc
Posts: 8614
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Doc » Sat Apr 28, 2018 2:12 pm

Kevin M wrote:
Sat Apr 28, 2018 1:56 pm
The Fidelity feature to download search results into a spreadsheet makes finding the good CD deals much more effective (and enabled me to produce the charts in this thread).
I just looked at Schwab and Vanguard. You can just copy and paste the search results into a spreadsheet. (You may have to do a page at a time depending on your search criteria but it's doable. At Vangaurd you can have 250 items per page so not a big deal.)
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.

User avatar
Kevin M
Posts: 10104
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: Yield Curves-Treasury, CD, AA/AAA muni

Post by Kevin M » Sat Apr 28, 2018 2:28 pm

Doc wrote:
Sat Apr 28, 2018 11:07 am
In my opinion all this makes questionable Kevin's assumption of holding to maturity as being the default case. Holding to maturity negates the benefits for "riding the yield" curve no matter how difficult it is hard it is to predict the actual value of the effect.
First a brief clarification. Holding to maturity is not an assumption, it's just what I plan to do with brokered CDs and municipal bonds, since the bid/ask spreads are too high to make selling before maturity a likely effective strategy. You probably would want to use Treasuries if your strategy involves selling before maturity (or a good direct CD with a reasonable early withdrawal penalty).

Of course we've had this discussion at length in other threads, but here I'll just mention that "riding the yield curve" only works if the yield curve doesn't move against you. Recently the yield curve has been moving against us, and "riding the yield curve" has had a negative contribution to returns, as opposed to the positive expected return if you assume a static yield curve. If this weren't the case, we'd see higher 1-year and 5-year returns for the Vanguard Treasury funds, for example.

I prefer to model this in terms of capital return and income return components for bonds bought at par. For example, with the current Treasury constant maturity yield curve (Treasury.gov), the 5-year Treasury yield is 2.77% and the 4-year Treasury yield is 2.68%. Assuming purchase at par, this gives a 1-year income return of 2.77% (the coupon rate), and an expected capital return component of 0.36%, assuming a static yield curve (the 0.36% comes from the value of a 4-year bond with yield of 2.68% and coupon of 2.77%).

However, if the 4-year yield increases to 2.77% in one year, your capital return component is 0%, and if it increases above that, your capital return component is negative. So you only have 9 basis points (bps) of buffer before your capital return goes negative. This is why a steeper yield curve provides higher probability of a positive capital return component. I don't think there's much "riding the yield curve" expected return built into the current, relatively flat yield curve.

Kevin
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

Post Reply