US T bills - first time user

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
geospatial
Posts: 62
Joined: Wed Jun 28, 2017 6:23 pm

Re: US T bills - first time user

Post by geospatial » Fri May 25, 2018 1:40 pm

gweezer wrote:
Fri May 25, 2018 12:20 pm
Thanks again Kevin. As the final chapter to this, what yield would I actually earn on this transaction? Using my numbers (even though it is now a hypothetical example), on 5/31/2019 I will have effectively received 3 coupon payments of $75 + $10K return of principal equals $10,225. I paid out $9995.53 initially, so I end up with $229.47 "profit". The yield calculation is (229.47/9995.53)*100 = 2.29%. To get the TEY (I'm a NY resident in the 24/6.65 brackets, no itemize): TEY = 2.29 (1 - .24) / (1 - .24 - .0665) which results in 2.509...does this look correct?
US treasuries are only state/local tax exempt, not federal, so I think in your calculation of TEY you need to exclude the 24% marginal rate.

FWIW, this thread has been fantastically educational and is exactly the reason I come to this forum. :sharebeer

pascalwager
Posts: 1216
Joined: Mon Oct 31, 2011 8:36 pm

Re: US T bills - first time user

Post by pascalwager » Fri May 25, 2018 6:07 pm

fortfun wrote:
Thu May 24, 2018 9:32 pm
livesoft wrote:
Thu May 24, 2018 9:20 pm
fortfun wrote:
Thu May 24, 2018 9:11 pm
Thanks Livesoft. Why did they return 8% back in the day that Your Money or Your Life was written? What happened to them?
The rate of inflation was crushed since then.
In her updated version of Your Life or Your Money, Robin recommends Index funds instead of t-bills. Will index funds go the way of t-bills? Fast forward 20 years, what will she recommend then, or are index funds here to stay?
Does she give any index fund examples? Vanguard, for example, has only the Treasury MM Fund ($50k min.) with a maturity of about seven weeks, compared to 13 weeks, 26 weeks, and one-year at Treasury Direct. Does the VG Treasury MM Fund meet all of her requirements even though it will give lower interest than long-term T-bills?

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Fri May 25, 2018 7:33 pm

gweezer wrote:
Fri May 25, 2018 12:20 pm
Thanks again Kevin. As the final chapter to this, what yield would I actually earn on this transaction? Using my numbers (even though it is now a hypothetical example), on 5/31/2019 I will have effectively received 3 coupon payments of $75 + $10K return of principal equals $10,225. I paid out $9995.53 initially, so I end up with $229.47 "profit". The yield calculation is (229.47/9995.53)*100 = 2.29%. To get the TEY (I'm a NY resident in the 24/6.65 brackets, no itemize): TEY = 2.29 (1 - .24) / (1 - .24 - .0665) which results in 2.509...does this look correct?
Without checking the math, the formulas look correct. The only caveat is that there is a time value of money component, so the earlier coupon payments (more generally, cash flows) have a larger present value than the later coupon payments. So your yield to maturity (or internal rate of return) should actually be slightly higher, but your calculation probably is close enough. You could use the XIRR spreadsheet function to get a more accurate internal rate of return.

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

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Fri May 25, 2018 7:38 pm

geospatial wrote:
Fri May 25, 2018 1:40 pm
gweezer wrote:
Fri May 25, 2018 12:20 pm
Thanks again Kevin. As the final chapter to this, what yield would I actually earn on this transaction? Using my numbers (even though it is now a hypothetical example), on 5/31/2019 I will have effectively received 3 coupon payments of $75 + $10K return of principal equals $10,225. I paid out $9995.53 initially, so I end up with $229.47 "profit". The yield calculation is (229.47/9995.53)*100 = 2.29%. To get the TEY (I'm a NY resident in the 24/6.65 brackets, no itemize): TEY = 2.29 (1 - .24) / (1 - .24 - .0665) which results in 2.509...does this look correct?
US treasuries are only state/local tax exempt, not federal, so I think in your calculation of TEY you need to exclude the 24% marginal rate.

FWIW, this thread has been fantastically educational and is exactly the reason I come to this forum. :sharebeer
Unless you itemize and get a full deduction for state income tax, the federal tax rate factors into TEY as shown in the equation. See this post for the derivations: viewtopic.php?f=10&t=248539. However, it only makes a difference of a few basis points relative to the simplified way you're thinking about, which applies if you itemize and get the full deduction for state income tax (which is going to be much rarer with tax changes in 2018).

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

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Fri May 25, 2018 9:16 pm

Kevin M wrote:
Fri May 25, 2018 7:33 pm
gweezer wrote:
Fri May 25, 2018 12:20 pm
Thanks again Kevin. As the final chapter to this, what yield would I actually earn on this transaction? Using my numbers (even though it is now a hypothetical example), on 5/31/2019 I will have effectively received 3 coupon payments of $75 + $10K return of principal equals $10,225. I paid out $9995.53 initially, so I end up with $229.47 "profit". The yield calculation is (229.47/9995.53)*100 = 2.29%. To get the TEY (I'm a NY resident in the 24/6.65 brackets, no itemize): TEY = 2.29 (1 - .24) / (1 - .24 - .0665) which results in 2.509...does this look correct?
Without checking the math, the formulas look correct. The only caveat is that there is a time value of money component, so the earlier coupon payments (more generally, cash flows) have a larger present value than the later coupon payments. So your yield to maturity (or internal rate of return) should actually be slightly higher, but your calculation probably is close enough. You could use the XIRR spreadsheet function to get a more accurate internal rate of return.

Kevin
I did the XIRR calculation, and was surprised that it came out to exactly your figure of 2.290%. Then I checked your math, and found that your equation result actually is 2.296%. So your math was slightly off. But I was wrong, in that the IRR is lower (not higher), because the more distant coupons have a smaller present value. Either way, your TEY is 2.51% or 2.52%, regardless of which way you calculate it.

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

User avatar
fortfun
Posts: 977
Joined: Tue Apr 19, 2016 7:31 pm

Re: US T bills - first time user

Post by fortfun » Fri May 25, 2018 10:09 pm

pascalwager wrote:
Fri May 25, 2018 6:07 pm
fortfun wrote:
Thu May 24, 2018 9:32 pm
livesoft wrote:
Thu May 24, 2018 9:20 pm
fortfun wrote:
Thu May 24, 2018 9:11 pm
Thanks Livesoft. Why did they return 8% back in the day that Your Money or Your Life was written? What happened to them?
The rate of inflation was crushed since then.
In her updated version of Your Life or Your Money, Robin recommends Index funds instead of t-bills. Will index funds go the way of t-bills? Fast forward 20 years, what will she recommend then, or are index funds here to stay?
Does she give any index fund examples? Vanguard, for example, has only the Treasury MM Fund ($50k min.) with a maturity of about seven weeks, compared to 13 weeks, 26 weeks, and one-year at Treasury Direct. Does the VG Treasury MM Fund meet all of her requirements even though it will give lower interest than long-term T-bills?
She doesn't give too much detail. Just says in the early days (80s and 90s, the first book) they recommended t-bills. Evidently they had a high return back then. She says that's no longer the case and recommends Vanguard Index (stocks and bonds). I need to re-read that last chapter. The premise of the book is basically to stop buying so much crap, put that money into index funds, and retire early. She spends a lot of time explaining how all of this consumerism is really bad for the environment too. Very good book. I recommend it to everyone.

geospatial
Posts: 62
Joined: Wed Jun 28, 2017 6:23 pm

Re: US T bills - first time user

Post by geospatial » Sat May 26, 2018 6:39 am

Kevin M wrote:
Fri May 25, 2018 7:38 pm
geospatial wrote:
Fri May 25, 2018 1:40 pm
gweezer wrote:
Fri May 25, 2018 12:20 pm
Thanks again Kevin. As the final chapter to this, what yield would I actually earn on this transaction? Using my numbers (even though it is now a hypothetical example), on 5/31/2019 I will have effectively received 3 coupon payments of $75 + $10K return of principal equals $10,225. I paid out $9995.53 initially, so I end up with $229.47 "profit". The yield calculation is (229.47/9995.53)*100 = 2.29%. To get the TEY (I'm a NY resident in the 24/6.65 brackets, no itemize): TEY = 2.29 (1 - .24) / (1 - .24 - .0665) which results in 2.509...does this look correct?
US treasuries are only state/local tax exempt, not federal, so I think in your calculation of TEY you need to exclude the 24% marginal rate.

FWIW, this thread has been fantastically educational and is exactly the reason I come to this forum. :sharebeer
Unless you itemize and get a full deduction for state income tax, the federal tax rate factors into TEY as shown in the equation. See this post for the derivations: viewtopic.php?f=10&t=248539. However, it only makes a difference of a few basis points relative to the simplified way you're thinking about, which applies if you itemize and get the full deduction for state income tax (which is going to be much rarer with tax changes in 2018).

Kevin
You're absolutely right, as usual. I got my wires crossed briefly by totally blanking on the fact that the federal tax rate was applied in the numerator as well. If I'm reading this correctly now, I think the translation is that a tax-free municipal bond with a yield of 2.29% is equivalent (produces the same net profit) to a taxable bond yielding 2.509% after the investor pays taxes on that at their marginal rates given in the example. Just seems weird to me because the 2.29% yield is coming from a profit that hasn't been taxed yet but will be, so wondering why calculating the TEY with that as an input is useful.

User avatar
jeffyscott
Posts: 7208
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

Re: US T bills - first time user

Post by jeffyscott » Sat May 26, 2018 7:21 am

Seemed like a strange calculation to me, too. But it works out, apply federal and state taxes to the 2.51% and fed only to 2.29% and you end up with the same after tax amount.

I believe it is because you are paying federal taxes on the portion of the fully taxable gains that goes to state taxes.
press on, regardless - John C. Bogle

geospatial
Posts: 62
Joined: Wed Jun 28, 2017 6:23 pm

Re: US T bills - first time user

Post by geospatial » Sat May 26, 2018 8:52 am

jeffyscott wrote:
Sat May 26, 2018 7:21 am
Seemed like a strange calculation to me, too. But it works out, apply federal and state taxes to the 2.51% and fed only to 2.29% and you end up with the same after tax amount.

I believe it is because you are paying federal taxes on the portion of the fully taxable gains that goes to state taxes.
When you explain it like that, the lightbulb goes off in my head and now it makes perfect sense. Thanks. I think part of the confusion arose because all the examples I've seen online are comparing tax-free vs. taxable (hence the name "Taxable Equivalent Yield") whereas this is an analysis of partly taxable vs. "totally" taxable relative to aggregate tax rates and their individual components, and could really be extended to any yield value where you want to compare different tax rates, even if one of them happens to be 0%.

BlackcatCA
Posts: 82
Joined: Sat Jan 20, 2018 2:55 pm

Re: US T bills - first time user

Post by BlackcatCA » Sat May 26, 2018 9:20 am

Kevin M wrote:
Thu May 24, 2018 12:07 pm

Finally, as already mentioned, you need to click the depth of book icon to find yields for the quantity you are interested in. So first pick maturity of interest, then click depth of book, then find highest yield with minimum quantity less than or equal to how many you want to buy. You can then click "buy" next to that one.

Kevin
Thank you everyone for the great educational posts on this topic. I just started dabbling in t- bills recently so this is very helpful.

I have a couple of questions for the bond experts:

1) All the listings have YTM yield displayed in the depth of books. Aren't those correct, or does one need to calculate each bond like the example?

2) in selecting bonds with the same maturity date, does it matter whether yield comes entirely from the bill ( e.g.zero coupon bill), or partly from the coupon? I have been avoiding coupons based on some older BH threads I've read. Maybe that's a mistake, and I should simply select the highest yield in the quantity I can afford to buy?

Thanks again.

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sat May 26, 2018 10:56 am

BlackcatCA wrote:
Sat May 26, 2018 9:20 am
1) All the listings have YTM yield displayed in the depth of books. Aren't those correct, or does one need to calculate each bond like the example?
The yields displayed are correct. Since there is no commission for Treasuries, you will see the same net yield displayed in the preview screen after you click buy and enter quantity, as long as you enter at least the minimum quantity for that offer, and as long as the yield/price hasn't changed since you were viewing the search results or depth of book screen.

The main confusion that started the discussion about different ways to calculate the yield were mostly related to accrued interest, and a misunderstanding about how bond prices are quoted.

A very fine point is that there are different ways to calculate yield, and you can get slightly different results depending on how you do it, but the differences mostly are in the third decimal place. But the yields you see displayed are good enough for making decisions.
2) in selecting bonds with the same maturity date, does it matter whether yield comes entirely from the bill ( e.g.zero coupon bill), or partly from the coupon? I have been avoiding coupons based on some older BH threads I've read. Maybe that's a mistake, and I should simply select the highest yield in the quantity I can afford to buy?
For bills, notes or bonds of same maturity, typically bonds with higher coupons will have lower yields than bonds with lower coupons (or no coupon)--at least if there is at least one coupon payment before maturity. This is because with higher coupons you get more of your money back sooner, and money you get sooner has a higher present value than the same amount received later. Another way to say this is that for a given maturity, a higher coupon bond has a lower duration, and bonds are priced more based on duration than on maturity.

So there's no particular reason to avoid coupon bonds, except that they may have a slightly lower yield at the specific maturity you're looking at. But that's only because they have less term risk; i.e., if rates increase, you can invest your earlier and/or larger coupon payments at the higher rates sooner.

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

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sat May 26, 2018 12:20 pm

Kevin M wrote:
Sat May 26, 2018 10:56 am
For bills, notes or bonds of same maturity, typically bonds with higher coupons will have lower yields than bonds with lower coupons (or no coupon)--at least if there is at least one coupon payment before maturity. This is because with higher coupons you get more of your money back sooner, and money you get sooner has a higher present value than the same amount received later. Another way to say this is that for a given maturity, a higher coupon bond has a lower duration, and bonds are priced more based on duration than on maturity.
We can see this effect by looking at a chart of yield vs. maturity for quotes pulled from Fidelity. However, there also are differences in yields for bonds of same maturity that aren't explained by the difference in coupon rates, which can also be seen by examining bonds at a maturity where we see significant differences in yields. Here is a chart of ask yields for bonds maturing in 3-5 years, just pulled from Fidelity.

Image

Usually, if not always, the sharp dips in yields are for Treasury bonds with high coupons, and the sharp peaks are for zero-coupon notes or bonds.

Looking at the first example at the far left of the chart for maturity 05/15/2021, the lowest yield is 2.56% for a bond with coupon 8.125%, and the highest yield of 2.62% is for a zero-coupon bond.

Maturity 11/15/2022 is an example where there's a higher peak but still with a significant dip below the curve. Here again the lowest yield of 2.70% is for a higher coupon bond, with coupon 7.625%, and the highest yield of 2.77% is for a zero-coupon bond.

However, there also is a zero-coupon bond with a yield of 2.71% at the same maturity. I see two differences between the two zero coupon bonds. One is that the for the lower yield the minimum quantity is 1, while for the higher yield it's 100--it is typical to see lower yields for smaller quantities. The other is that the lower yield is for a stripped principal bond while the higher yield is for a stripped interest bond.

Maturity 8/15/2022 is an example of significantly different zero-coupon yields where quantity doesn't explain it. The low yield of 2.67% still is for a high coupon bond with coupon of 7.250%, and the high yield of 2.75% is for a stripped interest zero-coupon bond. However, there is a stripped principal zero-coupon bond with same minimum quantity of 100 but yield of only 2.68%, so almost the same as the high coupon bond. I don't know enough about stripped interest vs. stripped coupon zeros to know if there is a rational explanation for this. Maybe someone else can explain it. Maybe there's a liquidity-based explanation.

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

User avatar
Doc
Posts: 8572
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: US T bills - first time user

Post by Doc » Sat May 26, 2018 12:34 pm

Kevin M wrote:
Sat May 26, 2018 12:20 pm
Here is a chart of ask yields for bonds maturing in 3-5 years, just pulled from Fidelity.
Looking at this type of detail during non-market hours is questionable. See if it holds up on Monday.
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: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sat May 26, 2018 12:57 pm

Doc wrote:
Sat May 26, 2018 12:34 pm
Kevin M wrote:
Sat May 26, 2018 12:20 pm
Here is a chart of ask yields for bonds maturing in 3-5 years, just pulled from Fidelity.
Looking at this type of detail during non-market hours is questionable. See if it holds up on Monday.
For the most part, it does--particularly the big dips for high coupon bonds. I've done the pulls when market is open, and see the same big dips for high coupon bonds. Haven't paid as much attention to zero-coupon bonds, nor the differences between stripped interest and stripped principal.

I have a chart for yields up to 3-year maturity pulled Friday morning at 10:14 AM Eastern. The big dips are there, and spot-checking them, they are for high coupon bonds. The zero-coupon peaks aren't evident, but neither are they in 0-3 year chart of yields pulled today. That seems to be a phenomenon at maturities beyond three years.

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

User avatar
Doc
Posts: 8572
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: US T bills - first time user

Post by Doc » Sat May 26, 2018 1:06 pm

Kevin M wrote:
Sat May 26, 2018 12:57 pm
For the most part, it does--particularly the big dips for high coupon bonds. I've done the pulls when market is open, and see the same big dips for high coupon bonds.
Are you able to see a distinction between a duration effect and a liquidity effect? I'm thinking that high coupon might mean a bond issued a long time ago and there may not be much of a market.

IIRC when we could get the yield charts at Fidelity there was usually a dip in yield for the on the run issue presumably because of high volume. Going the other way low volume could imply a higher spread and therefore maybe a higher yield.

Does any of this really matter? :happy
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: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sat May 26, 2018 3:17 pm

Doc wrote:
Sat May 26, 2018 1:06 pm
Kevin M wrote:
Sat May 26, 2018 12:57 pm
For the most part, it does--particularly the big dips for high coupon bonds. I've done the pulls when market is open, and see the same big dips for high coupon bonds.
Are you able to see a distinction between a duration effect and a liquidity effect? I'm thinking that high coupon might mean a bond issued a long time ago and there may not be much of a market.
Good question. But wouldn't lower liquidity result in higher yields instead of lower yields?

Lower liquidity -> lower demand -> lower price -> higher yield.

But we see lower yields in the higher coupon bonds. If lower liquidity was an explanation, we'd expect to see higher yields.
IIRC when we could get the yield charts at Fidelity there was usually a dip in yield for the on the run issue presumably because of high volume.

We can construct yield charts ourselves, and in one I shared last week, we did indeed see a lower yield for the on-the-run 10-year Treasury than for Treasuries of shorter maturity (and shorter duration!). Bonds/notes with maturities between 8.2 and 9.7 years had higher yields than the most recently auctioned 10-year note, even though the durations of these bonds were all less than the 10-year (however, I filtered out bonds with coupons > 6%, so no big dips in that chart). This is consistent with liquidity preference:

Higher liquidity -> higher demand -> higher price -> lower yield.
Going the other way low volume could imply a higher spread and therefore maybe a higher yield.
Yes, but we see lower yields for higher coupon bonds.

Interestingly, looking at bid/ask spreads for the bonds/notes maturing 8/15/2022, mentioned in my earlier reply with the chart, the 7.25% coupon bond has a bid/ask spread of 3.2 basis points, yet the yield is lower than the 1.625% coupon note with a spread of 1.3 bps. So just looking at these two, duration seems to explain the lower yield of the higher coupon bond.

However, looking at the zeros at this maturity, the higher-yield stripped interest bond does have a higher bid/ask spread at 3.8 bps, compared to the lower interest stripped coupon bond with spread of 3.3 bps. So liquidity could be an explanation here, but I think we'd have to look at a bigger data set to verify it.
Does any of this really matter? :happy
I think it matters if you want to understand why you might be getting higher (or lower) yield at a particular maturity. This sub-topic was initiated by this question:
2) in selecting bonds with the same maturity date, does it matter whether yield comes entirely from the bill ( e.g.zero coupon bill), or partly from the coupon? I have been avoiding coupons based on some older BH threads I've read. Maybe that's a mistake, and I should simply select the highest yield in the quantity I can afford to buy?
If you are selecting the highest yield at a given maturity, you may want to understand why that yield is higher than others for same maturity. In theory, duration should make a difference, as duration is a better measure of term risk than is maturity, and more term risk should be compensated with higher yield.

If you plan to hold to maturity, then liquidity shouldn't matter to you, and if there is an illiquidity yield premium, why not take it? It may matter more to you if you plan to sell before maturity, but the on the run Treasury you buy today with a liquidity price premium and yield discount will no longer be on the run when you sell it. So assuming a lower bid/ask for on the run, will the bid/ask still be lower when you sell it?

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

User avatar
Doc
Posts: 8572
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: US T bills - first time user

Post by Doc » Sat May 26, 2018 5:01 pm

Kevin M wrote:
Sat May 26, 2018 3:17 pm
But wouldn't lower liquidity result in higher yields instead of lower yields?
It could go either way. If you want to buy US Treasury 6.75% 08/15/2026 which has a dated date of 8/15/96 and there are very few available because they are all held by insurance companies to meet 30 year obligations when they were originally bought you would likely have to pay more for this than the on the run security. That is you will get a lower yield. Even without the liquidity issue you should expect a lower yield than the "on the run" security because of the shorter duration. Using the US Treasury 1.5% 08/15/2026 as the on the run security you indeed get a lower yield with the old bond. Furthermore the yield spread on the old note is 0.044 and the spread for the "on the run" is only 0.009. Again demonstrating liquidity issue.

The time for both of these securities to maturity is the same so if you are plotting YTM vs. time to maturity you should expect a "spikey" graph.
Kevin M wrote:
Sat May 26, 2018 3:17 pm
But we see lower yields in the higher coupon bonds. If lower liquidity was an explanation, we'd expect to see higher yields.
They should have lower yields because they have a shorter duration. The liquidity could go either way. Maybe Aetna has to dump a whole bunch of bonds they bought 20 years ago for some reason and there is nobody but Kevin, Doc and a few other Bogleheads to buy them. In that case we are going to get a higher yield. :)
Kevin M wrote:
Sat May 26, 2018 3:17 pm
If you are selecting the highest yield at a given maturity, you may want to understand why that yield is higher than others for same maturity. In theory, duration should make a difference, as duration is a better measure of term risk than is maturity, and more term risk should be compensated with higher yield.
Yep, part of the difference is in the duration and some may be due to liquidity. To get at the liquidity we would have to look at the Market Depth data and we'll have to wait until Tuesday to get meaningful numbers.

Maybe the bond traders can easily phantom all this but I certainly can't.

(And I used an "eight" for the "on the run issue" because it was at the top of the screen. :wink: )

Have a good holiday.
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.

Grt2bOutdoors
Posts: 19149
Joined: Thu Apr 05, 2007 8:20 pm
Location: New York

Re: US T bills - first time user

Post by Grt2bOutdoors » Sat May 26, 2018 5:50 pm

What happens if you bid for bill, it gets accepted but funds are not in account at time of settlement? Do the Tbill police come?
"One should invest based on their need, ability and willingness to take risk - Larry Swedroe" Asking Portfolio Questions

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sat May 26, 2018 9:19 pm

Doc wrote:
Sat May 26, 2018 5:01 pm
Kevin M wrote:
Sat May 26, 2018 3:17 pm
But wouldn't lower liquidity result in higher yields instead of lower yields?
It could go either way. If you want to buy US Treasury 6.75% 08/15/2026 which has a dated date of 8/15/96 and there are very few available because they are all held by insurance companies to meet 30 year obligations when they were originally bought you would likely have to pay more for this than the on the run security. That is you will get a lower yield.
OK, I could see this. So for some reason there is higher demand than supply, pushing price up and yield down. My question is why someone would want to buy this particular bond if there are alternatives with less demand but same maturity or duration, and higher yield.

And if it could go either way, it's not a good explanation for the consistently lower yields for the higher coupon bonds, but duration is a good explanation for this consistency.
Even without the liquidity issue you should expect a lower yield than the "on the run" security because of the shorter duration.
Exactly. That's the point I've been making. The consistently lower yields of the higher coupon bonds can be explained with duration only, without having to resort to liquidity as an explanation. So if liquidity could "go either way", it's not a good explanation for the consistently lower yields of the higher coupon bonds.
Using the US Treasury 1.5% 08/15/2026 as the on the run security you indeed get a lower yield with the old bond.
Of course. The 6.75% bond has a duration of 6.6 years vs. 7.7 years for the 1.5% note, so again, duration alone can explain this.
Furthermore the yield spread on the old note is 0.044 and the spread for the "on the run" is only 0.009. Again demonstrating liquidity issue.
I see 1.1 bps for high coupon vs 0.4 bps for lower coupon, but the conclusion is the same if using bid/ask as the proxy for liquidity (there are others). But as you say, liquidity could go either way, so we don't really know how liquidity is affecting the difference in yields, while we do know that duration is a technically valid explanation.
The time for both of these securities to maturity is the same so if you are plotting YTM vs. time to maturity you should expect a "spikey" graph.
Exactly, and the downward spikes for the high coupon bonds can be explained mostly by duration.

If I plot yield vs. duration, I still get a very spiky chart, but in this case it becomes very evident that the spikiness is due to the difference in yields between stripped interest and stripped principal zero-coupon bonds. There still might be some small dips for higher coupon bonds, but much less so than if you plot yield vs. maturity. So for me, the difference in yields between stripped interest and stripped principal zeros is the bigger mystery.

I haven't analyzed this to see if one is consistently higher than the other, but I did read that the supply of stripped principal zeros is much larger than the stripped interest zeros in dollar terms. An example was $16 billion stripped principal compared to $250 million stripped interest for a particular 30-year bond.

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

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sat May 26, 2018 9:30 pm

Grt2bOutdoors wrote:
Sat May 26, 2018 5:50 pm
What happens if you bid for bill, it gets accepted but funds are not in account at time of settlement? Do the Tbill police come?
You may not be allowed to even place such an order if you don't have sufficient assets in your account to cover it. You can get away with more in a margin account, but there still are going to be limits that protect the broker.

If you don't have sufficient cash in your settlement or core funds at settlement, the broker probably will liquidate some of your other holdings to pay for the trade. They probably will also then restrict your account to disallow placing trades without sufficient cash in your settlement fund(s).

At Fidelity, all money market funds are considered core funds, so if there is not enough in your settlement/cash fund, they will pull the rest from another money market fund. I just had that happen, as I was $0.01 short in my settlement fund at settlement, so they pulled it from a higher yielding money market fund, SPRXX, that I use to hold cash I'm gradually deploying into CDs or Treasuries. I usually transfer from SPRXX to the settlement fund before settlement, but in this case I guess I miscalculated by $0.01. I saw this ahead of time, but decided to let it ride to verify that they would pull the extra from SPRXX, as another forum member had said they would. I did not get any warnings. Fidelity shows cash available to trade as the sum of what is in the settlement fund and what is in SPRXX.

Vanguard does not seem to work this way, so you have to make sure to transfer from whatever other money market fund you might be using for higher yield into the lower yielding government money market fund that serves as the settlement fund before settlement.

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

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sat May 26, 2018 10:06 pm

Kevin M wrote:
Sat May 26, 2018 9:19 pm
If I plot yield vs. duration, I still get a very spiky chart, but in this case it becomes very evident that the spikiness is due to the difference in yields between stripped interest and stripped principal zero-coupon bonds. There still might be some small dips for higher coupon bonds, but much less so than if you plot yield vs. maturity. So for me, the difference in yields between stripped interest and stripped principal zeros is the bigger mystery.

I haven't analyzed this to see if one is consistently higher than the other, but I did read that the supply of stripped principal zeros is much larger than the stripped interest zeros in dollar terms. An example was $16 billion stripped principal compared to $250 million stripped interest for a particular 30-year bond.
So I did some analysis, and the results are kind of interesting. For shorter maturity STRIPS (zero-coupon bonds), stripped principal securities tend to have higher yields than stripped interest, but at maturity of 2/15/2020 this pattern reverses, and stripped interest securities have higher yields than stripped principal. The chart below shows this for maturities out to 10 years, with stripped interest labeled "I" and stripped principal labeled "P".

Image

Some of the labels are hard or impossible to read, but there are enough that are legible to discern that P is above I for the shorter maturities and I is above P for the longer maturities.

I do not know why.

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

Grt2bOutdoors
Posts: 19149
Joined: Thu Apr 05, 2007 8:20 pm
Location: New York

Re: US T bills - first time user

Post by Grt2bOutdoors » Sat May 26, 2018 11:30 pm

Kevin M wrote:
Sat May 26, 2018 9:30 pm
Grt2bOutdoors wrote:
Sat May 26, 2018 5:50 pm
What happens if you bid for bill, it gets accepted but funds are not in account at time of settlement? Do the Tbill police come?
You may not be allowed to even place such an order if you don't have sufficient assets in your account to cover it. You can get away with more in a margin account, but there still are going to be limits that protect the broker.

If you don't have sufficient cash in your settlement or core funds at settlement, the broker probably will liquidate some of your other holdings to pay for the trade. They probably will also then restrict your account to disallow placing trades without sufficient cash in your settlement fund(s).

At Fidelity, all money market funds are considered core funds, so if there is not enough in your settlement/cash fund, they will pull the rest from another money market fund. I just had that happen, as I was $0.01 short in my settlement fund at settlement, so they pulled it from a higher yielding money market fund, SPRXX, that I use to hold cash I'm gradually deploying into CDs or Treasuries. I usually transfer from SPRXX to the settlement fund before settlement, but in this case I guess I miscalculated by $0.01. I saw this ahead of time, but decided to let it ride to verify that they would pull the extra from SPRXX, as another forum member had said they would. I did not get any warnings. Fidelity shows cash available to trade as the sum of what is in the settlement fund and what is in SPRXX.

Vanguard does not seem to work this way, so you have to make sure to transfer from whatever other money market fund you might be using for higher yield into the lower yielding government money market fund that serves as the settlement fund before settlement.

Kevin
Thanks, I was thinking about using Treasury Direct.
"One should invest based on their need, ability and willingness to take risk - Larry Swedroe" Asking Portfolio Questions

User avatar
jeffyscott
Posts: 7208
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

Re: US T bills - first time user

Post by jeffyscott » Sun May 27, 2018 6:27 am

https://www.treasurydirect.gov/news/new ... marketable
On the issue date, the designated financial institution account or Zero-Percent C of I is debited for the purchase price (including any premium and/or accrued interest). Purchase requests are rejected if funds are insufficient to cover the cost of the securities.
I would assume if funds are coming from your checking account, it is also possible that the bank could allow an overdraft.
press on, regardless - John C. Bogle

MnD
Posts: 3718
Joined: Mon Jan 14, 2008 12:41 pm

Re: US T bills - first time user

Post by MnD » Sun May 27, 2018 8:01 am

After the June fed meeting and press conference I plan to start buying 90-day bills through Schwab at auction monthly with 1/4th of what I now have in money market. Goal is 25% in Schwab MM, 25% 1-30 days to maturity, 25% 31-60 days to maturity and 25% 61-90 days out. With one transaction of the same exact bill every month required, I like the simplicity. I'm not usually a cash guy but we have a fair amount saved up for a number of major purchases over the next 3-24 months. And I like to tinker and learn things.

As an aside, i was looking at the dates of the fed funds rate increases for 2017 and 2018 and daily t-bill yields.
Compared to 2017, the 30-day and 90-day T-bill rates in 2018 seem more favorable (higher) relative to the corresponding fed funds rate.
Not sure what's up with that - higher inflation, greater supply and possibility of four rate increases in 2018 I guess?

gmaynardkrebs
Posts: 886
Joined: Sun Feb 10, 2008 11:48 am

Re: US T bills - first time user

Post by gmaynardkrebs » Sun May 27, 2018 8:41 am

MnD wrote:
Sun May 27, 2018 8:01 am
After the June fed meeting and press conference I plan to start buying 90-day bills through Schwab at auction monthly with 1/4th of what I now have in money market. Goal is 25% in Schwab MM, 25% 1-30 days to maturity, 25% 31-60 days to maturity and 25% 61-90 days out. With one transaction of the same exact bill every month required, I like the simplicity. I'm not usually a cash guy but we have a fair amount saved up for a number of major purchases over the next 3-24 months. And I like to tinker and learn things.
Interesting, as my thinking is along the same lines. It's a pleasure to see rates that high once again. However, as I live in a high tax area (DC), the SEC yields on the Vanguard muni mm fund, and the Vanguard short muni VWSTX (avg duratio1.1 years), 1.24% and 1.74% respectively, seem like a better deal for me. I like VWSTX because I feel I'm being adequately rewarded for taking a rather small duration risk.

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sun May 27, 2018 10:45 am

MnD wrote:
Sun May 27, 2018 8:01 am
After the June fed meeting and press conference I plan to start buying 90-day bills through Schwab at auction monthly with 1/4th of what I now have in money market. Goal is 25% in Schwab MM, 25% 1-30 days to maturity, 25% 31-60 days to maturity and 25% 61-90 days out. With one transaction of the same exact bill every month required, I like the simplicity. I'm not usually a cash guy but we have a fair amount saved up for a number of major purchases over the next 3-24 months. And I like to tinker and learn things.
You might consider matching the maturity of this portion of your fixed income to the timing of your planned purchases, and possibly using CDs beyond about 18-month maturity, depending on your state income tax rate. The Treasury yield curve is quite steep out to 1-year maturity, and the CD-yield curve is very steep between 1-year and 2-year maturity (with a very a steep ramp between 1.9-year and 2.0-year maturity), so there's a good yield premium for taking the little bit of extra term risk. You also might skip the 1-month maturity Treasury at about 1.7%, as it's less than the Vanguard Prime MM yield of 1.88%, and even the 3-month is only about the same as Prime MM (don't know what Schwab MM is paying), but again, this depends on your marginal state income tax rate.

Here's the way it looks if you pay no state income tax:

Image

The 0-year point on the left vertical axis is the Prime MM yield of 1.88%. Note that the Treasury yield curve is flat to 3-month maturity. The 1-month is not shown, but if it were, we would see a downward slope in the Treasury yield curve from 0-year (Prime MM) to 1-month.

However, at my marginal tax rates of 27% federal and 8% state, Treasuries look better on a taxable-equivalent basis, as shown here:

Image

Still not showing the 1-month (as this data is pulled from the Fidelity yield summary page, and they don't include 1-month), but here we see that the 3-month taxable-equivalent yield (TEY) provides a nice yield premium over the MM fund--about 97 basis points/year, which is extremely steep. The 1-year Treasury at 2.58% TEY provides a yield premium of 70 bps/year compared to Prime MM, which still is very steep.

The curve flattens out quite a bit beyond 1-year, with only 21 bps of extra TEY for 2-year compared to 1-year, so if concerned about rising rates, you might want to limit your ladder to 1-year maturity if your tax rates were similar to mine.
As an aside, i was looking at the dates of the fed funds rate increases for 2017 and 2018 and daily t-bill yields.
Compared to 2017, the 30-day and 90-day T-bill rates in 2018 seem more favorable (higher) relative to the corresponding fed funds rate.
Not sure what's up with that - higher inflation, greater supply and possibility of four rate increases in 2018 I guess?
Astute observation. Here is a chart of the effective federal funds rate (EFFR) and 1-month and 3-month T-bills for the last three years, using 1-week averages to smooth the lines out a bit:

Image

We see that both T-bill yields were mostly below the EFFR before 2018, but now the 1-month is about the same and the 3-month is higher, and continuing to climb.

As somewhat of an aside, the Fed targets a lower and upper limit for the range of the FFR, and the EFFR is managed to stay within that range. Here is a chart showing the range and the daily EFFR for the last three years:

Image

One observation is that prior to the most recent increase in target rate range, there were sharp dips in EFFR on the last day of each month. Since the most recent increase, we don't see that. I have no idea why this is.

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

User avatar
Doc
Posts: 8572
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: US T bills - first time user

Post by Doc » Sun May 27, 2018 11:19 am

Kevin M wrote:
Sat May 26, 2018 9:19 pm
OK, I could see this. So for some reason there is higher demand than supply, pushing price up and yield down. My question is why someone would want to buy this particular bond if there are alternatives with less demand but same maturity or duration, and higher yield
One reason would be to establish a consistent guaranteed cash flow. Yield isn't the top priority in all cases.
Kevin M wrote:
Sat May 26, 2018 9:19 pm
I see 1.1 bps for high coupon vs 0.4 bps for lower coupon, but the conclusion is the same if using bid/ask as the proxy for liquidity (there are others).
I'm using a different data set and it was Saturday so I don't really belive any of the numbers are all that useful for these types of differences.
Kevin M wrote:
Sat May 26, 2018 9:19 pm
Exactly, and the downward spikes for the high coupon bonds can be explained mostly by duration.
I thought the same thing. But what's with the downward spikes on the daily Federal Funds Rate chart? I saw this a few days ago and have been scratching my head since. (But not very hard. :) )

Regarding liquidity: When markets are open it might be interesting to look at some market depth data and see if there is a significant liquidity difference between the "old" bonds and on the run issues. I looked yesterday and the market depth only was three or four where I usually see many more than that during market hours so I didn't pursue the idea.
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: 8572
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: US T bills - first time user

Post by Doc » Sun May 27, 2018 11:32 am

Grt2bOutdoors wrote:
Sat May 26, 2018 5:50 pm
What happens if you bid for bill, it gets accepted but funds are not in account at time of settlement? Do the Tbill police come?
When I buy anything that I don't know what the price is going to be before execution I virtually never have assets on hand before placing the order. Unless you were bad in the past and got yourself a "cash up front" restriction on your account you should have no problem at any broker. I do not know about Treasury direct. Edit: Oops I didn't catch the "settlement". See below.

I only had one problem several years ago while rolling a maturing note but that was caused by several interim trades between the auction and settlement date that messed up the broker's trade violation algorithm. The trade still went through but I got a trade violation that took a few days to remove. And it only took that long because the rep that was handling it was going on a long weekend and I told him it could wait until he got back.

So in addition to have assets sold to cover the trade as Kevin previously noted you will likely get a trade violation as well.
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.

anoop
Posts: 690
Joined: Tue Mar 04, 2014 1:33 am

Re: US T bills - first time user

Post by anoop » Sun May 27, 2018 11:46 am

Looking4Answers wrote:
Wed May 23, 2018 8:36 am
This might merit its own post, but I have some of the same questions about T Bills. However, I wonder about using Merrill Edge. I have seen recommendations in the past for Fidelity for this, but have incentive to keep transactions at Merrill Edge for Preferred Rewards. Any pros/cons to using Merrill Edge for T-bills?
Merrill Edge charges a hefty commission which is hidden in the (lower) rate which you receive.
viewtopic.php?f=1&t=237602&p=3714974&hi ... d#p3714911

I evaluated and decided to stay away from them.

In general, I prefer to do as little business as possible with the big banks. They have no ethics and have paid 100's of billions of dollars in fines.
https://www.marketwatch.com/story/banks ... 2018-02-20
Last edited by anoop on Sun May 27, 2018 12:38 pm, edited 2 times in total.

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sun May 27, 2018 12:29 pm

Kevin M wrote:
Sat May 26, 2018 9:19 pm
If I plot yield vs. duration, I still get a very spiky chart, but in this case it becomes very evident that the spikiness is due to the difference in yields between stripped interest and stripped principal zero-coupon bonds. There still might be some small dips for higher coupon bonds, but much less so than if you plot yield vs. maturity. <snip>
To investigate the duration impact for the large-coupon bonds, I filtered out the STRIPS (zero-coupon notes and bonds) for 7-10 year maturities, then charted yield vs. maturity date and yield vs. duration. First, the yield vs. maturity chart.

Image

We still see the large yield dips for the large-coupon bonds, but without the STRIPS, we don't see the large peaks.

Now the chart of yield vs. duration:

Image

We still see dips for larger-coupon bonds, but the curve is much smoother than for yield vs. maturity. The largest dip in the yield vs. duration chart is 2 basis points, while in the yield vs. maturity chart there are dips as large as 5 or 6 basis points, and a number at about 4 bps (in each chart, the minor vertical gridlines are 1 basis point).

Even in the yield vs. duration chart, the dips are indeed for the larger coupon bonds. Looking at the two dips of 2 basis points, the first is from 2.89% at 6.49 years (maturity = 05/15/2022, coupon = 2.125) to 2.87% at 6.58 years (08/15/2026, 6.750), and the second is from 2.90% at 6.89 years (11/15/2025, 2.250) to 2.88% at 6.93 years (02/15/2027, 6.625).

This seems to demonstrate that roughly 75% of the lower yields for higher-coupon bonds is explained by duration, but not all of it.

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

User avatar
Doc
Posts: 8572
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: US T bills - first time user

Post by Doc » Sun May 27, 2018 1:29 pm

Kevin M wrote:
Sun May 27, 2018 12:29 pm
We still see dips for larger-coupon bonds, but the curve is much smoother than for yield vs. maturity. The largest dip in the yield vs. duration chart is 2 basis points, while in the yield vs. maturity chart there are dips as large as 5 or 6 basis points, and a number at about 4 bps (in each chart, the minor vertical gridlines are 1 basis point).

Even in the yield vs. duration chart, the dips are indeed for the larger coupon bonds. Looking at the two dips of 2 basis points, the first is from 2.89% at 6.49 years (maturity = 05/15/2022, coupon = 2.125) to 2.87% at 6.58 years (08/15/2026, 6.750), and the second is from 2.90% at 6.89 years (11/15/2025, 2.250) to 2.88% at 6.93 years (02/15/2027, 6.625).

This seems to demonstrate that roughly 75% of the lower yields for higher-coupon bonds is explained by duration, but not all of it.

Kevin
How about doing this again when the market is open. Maybe do it on the bid instead of the ask to see if the dips become peaks indicating perhaps a spread/liquidity difference. :idea:
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.

MnD
Posts: 3718
Joined: Mon Jan 14, 2008 12:41 pm

Re: US T bills - first time user

Post by MnD » Sun May 27, 2018 1:41 pm

Kevin M wrote:
Sun May 27, 2018 10:45 am
MnD wrote:
Sun May 27, 2018 8:01 am
After the June fed meeting and press conference I plan to start buying 90-day bills through Schwab at auction monthly with 1/4th of what I now have in money market. Goal is 25% in Schwab MM, 25% 1-30 days to maturity, 25% 31-60 days to maturity and 25% 61-90 days out. With one transaction of the same exact bill every month required, I like the simplicity. I'm not usually a cash guy but we have a fair amount saved up for a number of major purchases over the next 3-24 months. And I like to tinker and learn things.
You might consider matching the maturity of this portion of your fixed income to the timing of your planned purchases, and possibly using CDs beyond about 18-month maturity, depending on your state income tax rate. The Treasury yield curve is quite steep out to 1-year maturity, and the CD-yield curve is very steep between 1-year and 2-year maturity (with a very a steep ramp between 1.9-year and 2.0-year maturity), so there's a good yield premium for taking the little bit of extra term risk. You also might skip the 1-month maturity Treasury at about 1.7%, as it's less than the Vanguard Prime MM yield of 1.88%, and even the 3-month is only about the same as Prime MM (don't know what Schwab MM is paying), but again, this depends on your marginal state income tax rate.
Thanks for your tips and graphs. I'm not going to buy anything shorter than 90-day but will buy one 90-day a month with 1/4th my cash until I have 1/4 in cash and 90-day t-bill ladder with 1/3 maturing every month. The timing of our purchases is really uncertain due to construction labor shortage issues delaying a remodel, when my retirement is processed and when an order for a small camper is made and filled. So I think I'll stay with 90-day bills and hopefully ride the yield increases up with each subsequent monthly purchase, assuming the 3-4 rate increase assumption holds. Schwab Prime MM is 1.74% so 90-day t-bill does buy a little more yield.

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sun May 27, 2018 1:45 pm

Doc wrote:
Sun May 27, 2018 1:29 pm
How about doing this again when the market is open. Maybe do it on the bid instead of the ask to see if the dips become peaks indicating perhaps a spread/liquidity difference. :idea:
That will be easy enough to do, but I don't think it will make much difference. We shall see.

The bid/ask spreads are indeed a bit larger for the high coupon bonds in this data set, but we're talking like 1.2 bps vs. 0.5 bps, so this would just make the dips a little smaller.

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

User avatar
Doc
Posts: 8572
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: US T bills - first time user

Post by Doc » Sun May 27, 2018 2:30 pm

Kevin M wrote:
Sun May 27, 2018 1:45 pm
Doc wrote:
Sun May 27, 2018 1:29 pm
How about doing this again when the market is open. Maybe do it on the bid instead of the ask to see if the dips become peaks indicating perhaps a spread/liquidity difference. :idea:
That will be easy enough to do, but I don't think it will make much difference. We shall see.

The bid/ask spreads are indeed a bit larger for the high coupon bonds in this data set, but we're talking like 1.2 bps vs. 0.5 bps, so this would just make the dips a little smaller.

Kevin
So that we are on the same page, are you using price spread or yield spreads?

I was seeing almost 4 bps in yield for the two 8/15/26 issues at Schwab yesterday. But I don't remember if I used qty 100 or the 25 default.

But that said, it's not likely to make much difference.

Were the spreads a lot more for the zeros you were looking at?
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
Blueskies123
Posts: 422
Joined: Sat Nov 15, 2014 7:18 pm
Location: South Florida

Re: US T bills - first time user

Post by Blueskies123 » Sun May 27, 2018 2:30 pm

Kevin, the spreadsheets you are posting are very useful. Can you explain how you do them? I assume you are downloading all the bond data points from Fidelity and then sorting them. Can you give us a few sorting tips or send a link so we can download them?

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sun May 27, 2018 2:36 pm

MnD wrote:
Sun May 27, 2018 1:41 pm
Thanks for your tips and graphs. I'm not going to buy anything shorter than 90-day but will buy one 90-day a month with 1/4th my cash until I have 1/4 in cash and 90-day t-bill ladder with 1/3 maturing every month. The timing of our purchases is really uncertain due to construction labor shortage issues delaying a remodel, when my retirement is processed and when an order for a small camper is made and filled. So I think I'll stay with 90-day bills and hopefully ride the yield increases up with each subsequent monthly purchase, assuming the 3-4 rate increase assumption holds. Schwab Prime MM is 1.74% so 90-day t-bill does buy a little more yield.
You seem committed to your plan, but here are a few more things to consider.

Since this is a taxable account, you should be making decisions based on taxable-equivalent yields (TEYs). I show how to calculate TEY in this post: Taxable Equivalent Yield (TEY). If you pay state income tax, your TEY for Treasuries is higher than the yield, because of the state income tax exemption.

You can earn 2.05% in a savings account with a minimum balance of $25K, with no term risk at all. So assuming no state income tax, you could just open a savings account at Northointe Bank to beat the 3-month yield, again, with no term risk. However, your TEY could be higher with the 3-month Treasury. You could at least park your cash there while building your ladder. That would get you an immediate increase of more than 30 basis points on your cash.

If the 3-month yield continues to rise at the same rate it has over the last six months, you would indeed do slightly better by rolling 3-month Treasuries for one year than in a 1-year Treasury held to maturity. The 3-month yield has risen from 0.49% 1.27% to 1.90% over the last six months, so 141 63 basis points, or an average of 70.5 about 32 basis points per quarter. Assuming no mistakes in my calculations, you would earn 2.99% 2.39% by rolling 3-month Treasuries for one year if they increased by 70.5 about 32 basis points per quarter. This of course is slightly more than the 1-year Treasury yield of 2.27%.

This applies only to the 1/4 of your cash that you will put into a 3-month Treasury now.

The institutional bond traders know more than we do, and if they thought the 3-month yield would increase at the indicated rate for the next year, the 1-year yield probably would be closer to 3% than 2.3%. On the other hand, even professionals are notoriously bad at predicting changes in interest rates over the next year, so maybe their apparent prediction of less rapid rate increases will turn out to be wrong.

Kevin
Last edited by Kevin M on Sun May 27, 2018 3:34 pm, edited 1 time in total.
Wiki ||.......|| Suggested format for Asking Portfolio Questions (edit original post)

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sun May 27, 2018 2:48 pm

Doc wrote:
Sun May 27, 2018 2:30 pm
So that we are on the same page, are you using price spread or yield spreads?
Since we're charting yields, I'm talking about yield spreads, since that's what would be relevant if I charted bid yield instead of ask yield.
I was seeing almost 4 bps in yield for the two 8/15/26 issues at Schwab yesterday. But I don't remember if I used qty 100 or the 25 default.
I'm using the default minimum quantity at Fidelity, which for these two issues is 200.
Were the spreads a lot more for the zeros you were looking at?
For this same maturity, the spread for the stripped principal issue is 3.7 bps, and for the stripped interest issue it's 3.3 bps. Minimum quantity for both is 100. So yes, somewhat more.

In general the spreads for the STRIPS are higher. The average yield bid/ask for the 0-10 STRIPS is 4.5 bps, and I see several higher than 10 bps, and one higher than 20 bps.

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

lack_ey
Posts: 6612
Joined: Wed Nov 19, 2014 11:55 pm

Re: US T bills - first time user

Post by lack_ey » Sun May 27, 2018 2:51 pm

Kevin M wrote:
Sun May 27, 2018 2:36 pm
If the 3-month yield continues to rise at the same rate it has over the last six months, you would indeed do better by rolling 3-month Treasuries for one year than in a 1-year Treasury held to maturity. The 3-month yield has risen from 0.49% to 1.90% over the last six months, so 141 basis points, or an average of 70.5 basis points per quarter. Assuming no mistakes in my calculations, you would earn 2.99% by rolling 3-month Treasuries for one year if they increased by 70.5 basis points per quarter. This of course is more than the 1-year Treasury yield of 2.27%.
0.49% to 1.90% in six months? You sure you're not looking at eighteen months?

For the 3-month CMT series I see 0.48% on 2016-11-28, 1.27% on 2017-11-27.

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sun May 27, 2018 3:19 pm

Blueskies123 wrote:
Sun May 27, 2018 2:30 pm
Kevin, the spreadsheets you are posting are very useful. Can you explain how you do them? I assume you are downloading all the bond data points from Fidelity and then sorting them. Can you give us a few sorting tips or send a link so we can download them?
Glad you find them useful.

I find them particularly useful in quickly identifying what looks good at a particular time when the market is open, more for CDs than for Treasuries, but also to compare them to each other. This also allows me to calculate TEY if I'm buying in a taxable account. When you chart the results, you can quickly identify outliers that are offering higher than average yields (again mostly useful for CDs as opposed to Treasuries).

My last raw download of CDs out to 3-year maturity included about 750 CDs, but then in a separate sheet, I do some additional filtering to screen out the ones that have lower yields, and ended up with only about 35 CDs to chart and take a closer look at. Similarly, for the charts in this post, I start with all the 0-10 year Treasuries, about 330 of them, then in separate sheets filter by maturity, or filter to just look at STRIPS or to filter out STRIPS, for example.

Although you can look at the Fidelity yield summary page without logging on, as far as I can tell, you must log on to your account to do searches. So I can't really send you a link to download them the way I do.

You can also get prices and yields from the WSJ Treasury Quotes page. You can copy/paste these into a spreadsheet. If you know how to use ImportHTML in Google Sheets, you can import them that way. With this URL in cell B1, http://online.wsj.com/mdc/public/page/2 ... nav_2_3020, use =importhtml(B1,"table",3). I haven't used this in awhile, since I'm usually interested in prices/yields for bonds I can buy when the market is open, and I'm just leveraging that for what I'm posting here.

If you have a Fidelity account, there is a download button on the Fidelity search results screen. You can sort by any of the columns in the search results screen, and I sort by maturity before downloading, so I don't have to bother doing that in the spreadsheet. It seems that your previous sort choice carries over from one search to the next, so by default they usually are sorted by maturity.

In my search criteria I limit to the maturities of interest. For my daily buying decisions, this is out to 3-year maturity, since yield curves are too flat beyond that to interest me. Obviously for the charts in this thread I'm going out to 10 years.

There are a number of other parameters you can use to restrict your search results, such as minimum yield, coupon rate, etc. I use these, depending on what my purpose is.

The downloaded data gives you everything you need to calculate duration except settlement date, so I enter the appropriate settlement date into a cell and use that in the DURATION function. You also can calculate maturity if you want to check whatever Fidelity shows. Also, for CDs and munis, for which there is a commission, I calculate net yield by adding 0.1 to the ask price, and using the YIELD function.

Hope that helps,

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

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sun May 27, 2018 3:25 pm

lack_ey wrote:
Sun May 27, 2018 2:51 pm
0.49% to 1.90% in six months? You sure you're not looking at eighteen months?

For the 3-month CMT series I see 0.48% on 2016-11-28, 1.27% on 2017-11-27.
Oops, you're right. I'll redo the calculations now.
:oops:

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

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sun May 27, 2018 3:35 pm

Kevin M wrote:
Sun May 27, 2018 3:25 pm
lack_ey wrote:
Sun May 27, 2018 2:51 pm
0.49% to 1.90% in six months? You sure you're not looking at eighteen months?

For the 3-month CMT series I see 0.48% on 2016-11-28, 1.27% on 2017-11-27.
Oops, you're right. I'll redo the calculations now.
:oops:

Kevin
Done: viewtopic.php?f=10&t=250059&p=3946494#p3946494.

Thanks for the catch, lack_ey.

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

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Tue May 29, 2018 10:46 am

Kevin M wrote:
Sun May 27, 2018 1:45 pm
Doc wrote:
Sun May 27, 2018 1:29 pm
How about doing this again when the market is open. Maybe do it on the bid instead of the ask to see if the dips become peaks indicating perhaps a spread/liquidity difference. :idea:
That will be easy enough to do, but I don't think it will make much difference. We shall see.

The bid/ask spreads are indeed a bit larger for the high coupon bonds in this data set, but we're talking like 1.2 bps vs. 0.5 bps, so this would just make the dips a little smaller.

Kevin
Here are the charts again with data downloaded at 10:53 AM Eastern this morning (market open). This time I've included both bid and ask yield.

Image

Image

We see basically the same thing as before. The largest dips on the order of 5-6 basis points if graphed by maturity date, but only 1-2 bps if graphed by duration. So shorter duration explains most, but not all, of the lower yields for the higher coupon bonds.

As predicted, the larger bid/ask spreads for higher coupon bonds just makes the dips for the bid yields slightly smaller.

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

User avatar
Doc
Posts: 8572
Joined: Sat Feb 24, 2007 1:10 pm
Location: Two left turns from Larry

Re: US T bills - first time user

Post by Doc » Tue May 29, 2018 11:34 am

Kevin M wrote:
Tue May 29, 2018 10:46 am
We see basically the same thing as before. The largest dips on the order of 5-6 basis points if graphed by maturity date, but only 1-2 bps if graphed by duration. So shorter duration explains most, but not all, of the lower yields for the higher coupon bonds.

As predicted, the larger bid/ask spreads for higher coupon bonds just makes the dips for the bid yields slightly smaller.
Thanks Kevin.
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.

bayview
Posts: 1550
Joined: Thu Aug 02, 2012 7:05 pm
Location: WNC

Re: US T bills - first time user

Post by bayview » Sat Jun 09, 2018 1:28 pm

A bump for a question from another first-timer involving discount rates vs YTM:

DH has brokered CD's (1-3 years) and T-bills (3, 6, and 12 months) in his IRA.

When the bond markets are open, I'm working. When they're closed, I can't see what's up with new issues.

I am quite maddened that Vanguard displays the coupon for the CDs, but not for the T-bills. Yes, I realize that it's because T-bills are bought at a discount, but I have a hard time comparing them. I didn't know this when I bought them, so now I'm scrambling a bit. I was able to get their discount rates from the confirmations.

Given these two things:
- no tax considerations, because these are held in an IRA, and
- they will be held to maturity,
how do I compare the quoted discount rates on the T-bills to the YTMs on the CDs? Will they in fact be the same, since the various investments will be held to maturity?? (I'm certain that's far too simple. :oops: )

If there is any easy way to run these numbers, ideally without a formula with a sigma symbol in there anywhere :P , I'd appreciate any guidance.
The continuous execution of a sound strategy gives you the benefit of the strategy. That's what it's all about. --Rick Ferri

User avatar
jeffyscott
Posts: 7208
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

Re: US T bills - first time user

Post by jeffyscott » Sat Jun 09, 2018 2:20 pm

Just bought my first one this week, for my purposes the "Investment/Interest Rate" shown on TD or the rates here: https://www.treasury.gov/resource-cente ... =billrates are as much as I need to know.

But since I was curious about how you would determine the YTM for security with less than a year to maturity, I did find this:
https://pocketsense.com/calculate-yield ... 59681.html

At another link it was indicated that the ask yield is the bond yield equivalent (b.y.e), FWIW.
press on, regardless - John C. Bogle

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sat Jun 09, 2018 2:23 pm

bayview wrote:
Sat Jun 09, 2018 1:28 pm
A bump for a question from another first-timer involving discount rates vs YTM:

DH has brokered CD's (1-3 years) and T-bills (3, 6, and 12 months) in his IRA.

When the bond markets are open, I'm working. When they're closed, I can't see what's up with new issues.
You can always see what's up with new issue CDs, so I'm puzzled by this statement. You can see them here at Vanguard, for example, without even logging on: https://personal.vanguard.com/us/FixedIncomeHome. If you click on say the 2.80% new-issue CD, you will see the various new issues available.

Are you talking about Treasuries at auction? You can see those too and enter an order as long as the auction is open. I am thinking of entering an order for a 6-month Treasury this weekend.
I am quite maddened that Vanguard displays the coupon for the CDs, but not for the T-bills. Yes, I realize that it's because T-bills are bought at a discount, but I have a hard time comparing them. I didn't know this when I bought them, so now I'm scrambling a bit. I was able to get their discount rates from the confirmations.
Vanguard displays the coupon rates for all Treasuries if you search on the secondary market. You don't see the coupon rate on the new issue search results screen, but you can see the coupon rate if you click buy to bring up the details of the specific Treasury (you can then click Cancel if you don't want to actually buy). But the coupon rate always is 0% for T-bills, so there's no need to even bother.
Given these two things:
- no tax considerations, because these are held in an IRA, and
- they will be held to maturity,
how do I compare the quoted discount rates on the T-bills to the YTMs on the CDs? Will they in fact be the same, since the various investments will be held to maturity?? (I'm certain that's far too simple. :oops: )
If buying at auction, Vanguard displays the "indicative yield", which is the current best guess for what the yield (to maturity) will be. I believe the yield to maturity also is shown on your confirmation for T-bills.

The yield to maturity also is shown on the auction results on Treasury Direct institutional: https://www.treasurydirect.gov/instit/a ... esults.htm. For example, at that site, I drill down on Bill -> 26 week. Most recent auction was 6/4/2018, so I click on the PDF for the competitive results for this one. I believe the relevant rate is the Investment Rate, since the footnote for this says it's the equivalent coupon-issue yield, so it's the yield you would use to compare to a CD or Treasury with a coupon, but others can chime in if this is incorrect. The investment rate for this bill is 2.121%.

For deciding between buying, say, a 6-month Treasury or 6-month CD in an IRA, I'd use the indicative yield for the Treasury of 2.114%, and the new-issue 6-month CD rate of 2.10%. Since the Treasury is slightly higher, and since Treasuries are more liquid (although if held to maturity this doesn't matter), I'd probably go with the Treasury. Of course you don't know exactly what yield you'll actually get for the Tbill, and since the yields are close, if you want to know for sure what yield you'll get, buy the CD.

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

bayview
Posts: 1550
Joined: Thu Aug 02, 2012 7:05 pm
Location: WNC

Re: US T bills - first time user

Post by bayview » Sat Jun 09, 2018 2:24 pm

jeffyscott wrote:
Sat Jun 09, 2018 2:20 pm

...But since I was curious about how you would determine the YTM for security with less than a year to maturity, I did find this:
https://pocketsense.com/calculate-yield ... 59681.html

At another link it was indicated that the ask yield is the bond yield equivalent (b.y.e), FWIW.
Thanks for the link. No sigma symbols! :sharebeer

From the link:
A debt security's "yield-to-maturity (YTM)" refers to how much of a return it will provide if held to maturity. However, YTM is usually calculated by the year. To calculate YTM for a security maturing in less than a year, you need to calculate the "Bond Yield Equivalent (BYE)."

Subtract the asking price of the T-bill from its par value, the dollar amount that the T-bill will be redeemable for at maturity.

Divide the answer from Step 1 by the par value.

Divide 182 days (the term of the T-bill) by 364 days (the number of days in the financial calendar year).

Divide the result from Step 2 by 0.5 (the result from Step 3).

Multiply by 100 percent to convert the BYE to a percentage. By substituting this value for YTM, you can better compare the attractiveness of a six-month T-bill compared to securities with multi-year terms.
There are some additional warnings on the link itself.
The continuous execution of a sound strategy gives you the benefit of the strategy. That's what it's all about. --Rick Ferri

bayview
Posts: 1550
Joined: Thu Aug 02, 2012 7:05 pm
Location: WNC

Re: US T bills - first time user

Post by bayview » Sat Jun 09, 2018 2:38 pm

Kevin M wrote:
Sat Jun 09, 2018 2:23 pm
bayview wrote:
Sat Jun 09, 2018 1:28 pm
A bump for a question from another first-timer involving discount rates vs YTM:

DH has brokered CD's (1-3 years) and T-bills (3, 6, and 12 months) in his IRA.

When the bond markets are open, I'm working. When they're closed, I can't see what's up with new issues.
You can always see what's up with new issue CDs, so I'm puzzled by this statement. You can see them here at Vanguard, for example, without even logging on: https://personal.vanguard.com/us/FixedIncomeHome. If you click on say the 2.80% new-issue CD, you will see the various new issues available.

Are you talking about Treasuries at auction? You can see those too and enter an order as long as the auction is open. I am thinking of entering an order for a 6-month Treasury this weekend.

<snip>

Kevin
Thanks, Kevin. I appreciate your reply, but my problem was that this info is often ephemeral. Although I was able to do all that and make comparisons when the Treasuries were available during auction, much of this vanishes afterwards. Similar issue when the bond markets are closed, which is the only time I can log on and mess around. (And yes, it was the new Treasuries, not new issue CDs, that I was ranting about.)

I was able to use the information at the time I was buying, but since different things were bought at different times, I'm now trying to figure out what (if anything) I did right, lol. btw, the recent run-up in CD's is doubtless due to the fact that we purchased them in early-to-mid April, and the coupon for each is now pretty much exactly that of the next-younger CD (the 3-year CD pays what the 2-years now pay, the 2-year now pays what the 18-months pay, etc.) You're welcome, everyone. :annoyed :D
If buying at auction, Vanguard displays the "indicative yield", which is the current best guess for what the yield (to maturity) will be. I believe the yield to maturity also is shown on your confirmation for T-bills.

The yield to maturity also is shown on the auction results on Treasury Direct institutional: https://www.treasurydirect.gov/instit/a ... esults.htm. For example, at that site, I drill down on Bill -> 26 week. Most recent auction was 6/4/2018, so I click on the PDF for the competitive results for this one. I believe the relevant rate is the Investment Rate, since the footnote for this says it's the equivalent coupon-issue yield, so it's the yield you would use to compare to a CD or Treasury with a coupon, but others can chime in if this is incorrect. The investment rate for this bill is 2.121%.

For deciding between buying, say, a 6-month Treasury or 6-month CD in an IRA, I'd use the indicative yield for the Treasury of 2.114%, and the new-issue 6-month CD rate of 2.10%. Since the Treasury is slightly higher, and since Treasuries are more liquid (although if held to maturity this doesn't matter), I'd probably go with the Treasury. Of course you don't know exactly what yield you'll actually get for the Tbill, and since the yields are close, if you want to know for sure what yield you'll get, buy the CD.

Kevin
Ah, this is great to know! It doesn't matter if it's not precise; just some sort of decent comparison. That will be very helpful as we continue to build the ladder.

edit to add:

I just saw this:
... I believe the yield to maturity also is shown on your confirmation for T-bills.
Alas, it is not; only the discount rate. This is a copy/paste from one of the confirmations (for the three-month). The formatting went to pieces, of course, but I didn't want to do an image:

U S TREASURY BILL DUE 07/12/18 DTD 01/11/18
Symbol:— CUSIP: 912796-PQ-6
04/09/2018 04/12/2018 Buy 10,000.00000 $99.56648 $9,956.65 — $9,956.65
(Notes: 1, 4) ——
1.715% DISCOUNT RATE
The continuous execution of a sound strategy gives you the benefit of the strategy. That's what it's all about. --Rick Ferri

User avatar
Kevin M
Posts: 10079
Joined: Mon Jun 29, 2009 3:24 pm
Contact:

Re: US T bills - first time user

Post by Kevin M » Sat Jun 09, 2018 2:55 pm

I see that you're right about the Tbill confirmation only showing the discount rate (I looked at one of mine). So what you can do is go to the TD site and look up the investment rate for that bill. For example, for the 3-month bill I bought maturing 6/21/2018, the discount rate on the confirmation is 1.779%. The auction result PDF shows the "high rate" as 1.780%, and I assume the difference is just a rounding thing. The investment rate shown in the TD PDF is 1.813%.

I also get 1.813% using the spreadsheet YIELD function with the frequency parameter set to 1 and day count convention set to 1 (actual/actual) or 3 (actual/365), or with the frequency parameter set to 2 and daycount convention set to 3. I would think frequency should be set to 2 for comparison to Treasury coupon bonds, since they pay interest semi-annually.

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

bayview
Posts: 1550
Joined: Thu Aug 02, 2012 7:05 pm
Location: WNC

Re: US T bills - first time user

Post by bayview » Sat Jun 09, 2018 3:02 pm

Kevin M wrote:
Sat Jun 09, 2018 2:55 pm
I see that you're right about the Tbill confirmation only showing the discount rate (I looked at one of mine). So what you can do is go to the TD site and look up the investment rate for that bill. For example, for the 3-month bill I bought maturing 6/21/2018, the discount rate on the confirmation is 1.779%. The auction result PDF shows the "high rate" as 1.780%, and I assume the difference is just a rounding thing. The investment rate shown in the TD PDF is 1.813%.

I also get 1.813% using the spreadsheet YIELD function with the frequency parameter set to 1 and day count convention set to 1 (actual/actual) or 3 (actual/365), or with the frequency parameter set to 2 and daycount convention set to 3. I would think frequency should be set to 2 for comparison to Treasury coupon bonds, since they pay interest semi-annually.

Kevin
Thanks! I'll try the TD site.

I did run the numbers on the formula that jeffyscott supplied and wound up with what appears to be exactly 1/2 of what it should be (and that's with dividing by 0.5.) I have decided that my brain objects to having to work on Saturday...
The continuous execution of a sound strategy gives you the benefit of the strategy. That's what it's all about. --Rick Ferri

Post Reply