CDs role in Fixed income part of portfolio

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
michaeljc70
Posts: 3746
Joined: Thu Oct 15, 2015 3:53 pm

CDs role in Fixed income part of portfolio

Post by michaeljc70 » Fri Sep 14, 2018 6:03 pm

I'm interested in what role CDs should play in a FI part of a portfolio, particularly near/in retirement compared to Total Bond or similar. Obviously, CDs don't lose principal value while they also don't gain outsized returns in volatile/dropping rate environments like longer term bonds (like during and after 2008). How do CDs compare long term to the performance of Total Bond (obviously duration of the CD comes into play)? Do you allocate a % of FI to CDs or X months of living expenses or????

I'd rather not fool with CDs/CD ladders, but am questioning that. My AA is 75/25 and I don't plan on changing that so maybe it is moot with only 25% FI.

Thesaints
Posts: 1619
Joined: Tue Jun 20, 2017 12:25 am

Re: CDs role in Fixed income part of portfolio

Post by Thesaints » Fri Sep 14, 2018 6:13 pm

Your bond component should be tuned to your needs. It does not have to be the same as the total bond market.

User avatar
Cyclesafe
Posts: 658
Joined: Wed Dec 31, 2014 1:03 pm

Re: CDs role in Fixed income part of portfolio

Post by Cyclesafe » Fri Sep 14, 2018 6:18 pm

CD's make sense for a retail investor who wants zero risk and accepts the lower return thus associated with them. Interim values of secondary CD's between purchase and maturity do fluctuate, however, just like bond funds. Sometimes buying CD's are better than buying similarly risk free treasury bonds (via Treasury Direct), and sometimes not - even when considering that states do not tax interest earned from treasuries.

The other options for fixed you mention are not risk free so they offer a premium return to compensate for this.

There's nothing whatsoever wrong with CD's being the sole investment in fixed. Many Bogleheads do this.
Last edited by Cyclesafe on Fri Sep 14, 2018 7:02 pm, edited 1 time in total.

Silk McCue
Posts: 1312
Joined: Thu Feb 25, 2016 7:11 pm

Re: CDs role in Fixed income part of portfolio

Post by Silk McCue » Fri Sep 14, 2018 6:25 pm

My wife Is just over 2 years from retirement at ages 60/62 with A COLA’d pension that provides a nice base for us. I just moved north of $200k to Achieva Credit Union to a 5 year IRA CD at 4.2% a couple of months back from my Vanguard Bond holdings. That rate has been available since May and is available today. Great early withdrawal terms for folks over 59.5 (2 years from now for me) - 0 penalty. With that CD and other pile of guaranteed funds we have we are coverered for retirement along with the Pension to meet base needs and lots of fun until age 68/70 when Social Security kicks in. Zero equities will need to be touched during that time except for Roth Conversions.

For me, going to a CD at the nations highest rate was an easy decision given our circumstances. It could be that my bond holdings might very well have outperformed in the same timeframe but I have even more peace of mind by knowing what bird I have in the hand.

Cheers

aristotelian
Posts: 4767
Joined: Wed Jan 11, 2017 8:05 pm

Re: CDs role in Fixed income part of portfolio

Post by aristotelian » Fri Sep 14, 2018 6:47 pm

Cyclesafe wrote:
Fri Sep 14, 2018 6:18 pm
CD's make sense for a retail investor who wants zero risk and accepts the lower return thus associated with them. Interim values of the CD's between purchase and maturity do fluctuate, however, just like bond funds. Sometimes buying CD's are better than buying similarly risk free treasury bonds (via Treasury Direct), and sometimes not - even when considering that states do not tax interest earned from treasuries.

The other options for fixed you mention are not risk free so they offer a premium return to compensate for this.

There's nothing whatsoever wrong with CD's being the sole investment in fixed. Many Bogleheads do this.
CDs do not necessarily have lower returns vs similar risk instruments. 5 year CDs typically beat Treasuries of the same duration. I would frame the choice as liquidity of Treasuries vs higher yield of CDs.

michaeljc70
Posts: 3746
Joined: Thu Oct 15, 2015 3:53 pm

Re: CDs role in Fixed income part of portfolio

Post by michaeljc70 » Fri Sep 14, 2018 7:05 pm

aristotelian wrote:
Fri Sep 14, 2018 6:47 pm
Cyclesafe wrote:
Fri Sep 14, 2018 6:18 pm
CD's make sense for a retail investor who wants zero risk and accepts the lower return thus associated with them. Interim values of the CD's between purchase and maturity do fluctuate, however, just like bond funds. Sometimes buying CD's are better than buying similarly risk free treasury bonds (via Treasury Direct), and sometimes not - even when considering that states do not tax interest earned from treasuries.

The other options for fixed you mention are not risk free so they offer a premium return to compensate for this.

There's nothing whatsoever wrong with CD's being the sole investment in fixed. Many Bogleheads do this.
CDs do not necessarily have lower returns vs similar risk instruments. 5 year CDs typically beat Treasuries of the same duration. I would frame the choice as liquidity of Treasuries vs higher yield of CDs.
I think a CD ladder could eliminate most liquidity issues (assuming expense are known). Also, although bond funds can drop in value, you don't hear much about people not selling them because the price is low like stocks. Drops tend to be minor.

aristotelian
Posts: 4767
Joined: Wed Jan 11, 2017 8:05 pm

Re: CDs role in Fixed income part of portfolio

Post by aristotelian » Fri Sep 14, 2018 7:58 pm

michaeljc70 wrote:
Fri Sep 14, 2018 7:05 pm

I think a CD ladder could eliminate most liquidity issues (assuming expense are known). Also, although bond funds can drop in value, you don't hear much about people not selling them because the price is low like stocks. Drops tend to be minor.
I am becoming a big fan of the CD ladder. Still, much tougher to liquidate than a bond fund which you can sell at the click of a button.

keepingitsimple
Posts: 143
Joined: Sun Mar 18, 2018 8:20 pm

Re: CDs role in Fixed income part of portfolio

Post by keepingitsimple » Fri Sep 14, 2018 8:09 pm

michaeljc70 wrote:
Fri Sep 14, 2018 6:03 pm
I'd rather not fool with CDs/CD ladders, but am questioning that.
I am a fan of CD's and CD ladders for fixed income. Additionally, I find CD ladders easy to maintain.

venkman
Posts: 708
Joined: Tue Mar 14, 2017 10:33 pm

Re: CDs role in Fixed income part of portfolio

Post by venkman » Fri Sep 14, 2018 9:05 pm

Right now, you can build a 5-year CD ladder with brokered CD's at Vanguard that yields 3.01% in the first year. If you could reinvest the maturing 1-year CD at current 5-year rates, the average yield goes up to 3.19% in year two. The average maturity of the ladder will be 2.5 years, and your principal is effectively as safe as it would be in Treasuries.

For comparison, VG's Short Term Treasury Index (VSBSX) has an average maturity of 2.0 years, and a current SEC yield of 2.58%.

If you have a significant amount in FI and don't need it all to be completely liquid, a CD ladder is a compelling possibility for at least part of your FI allocation, especially if the money isn't subject to state taxes. I'd still keep some assets in bond funds, just for easy rebalancing.

User avatar
munemaker
Posts: 3512
Joined: Sat Jan 18, 2014 6:14 pm

Re: CDs role in Fixed income part of portfolio

Post by munemaker » Sat Sep 15, 2018 8:12 am

aristotelian wrote:
Fri Sep 14, 2018 7:58 pm
michaeljc70 wrote:
Fri Sep 14, 2018 7:05 pm

I think a CD ladder could eliminate most liquidity issues (assuming expense are known). Also, although bond funds can drop in value, you don't hear much about people not selling them because the price is low like stocks. Drops tend to be minor.
I am becoming a big fan of the CD ladder. Still, much tougher to liquidate than a bond fund which you can sell at the click of a button.
If you buy brokered CDs or treasuries, you can also sell at the click of a button.

User avatar
munemaker
Posts: 3512
Joined: Sat Jan 18, 2014 6:14 pm

Re: CDs role in Fixed income part of portfolio

Post by munemaker » Sat Sep 15, 2018 8:15 am

venkman wrote:
Fri Sep 14, 2018 9:05 pm
VG's Short Term Treasury Index (VSBSX) has an average maturity of 2.0 years, and a current SEC yield of 2.58%.
Note: The YTD return of VSBSX is 0.10%. The one year total return is -0.31%.

User avatar
jeffyscott
Posts: 7311
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

Re: CDs role in Fixed income part of portfolio

Post by jeffyscott » Sat Sep 15, 2018 9:42 am

munemaker wrote:
Sat Sep 15, 2018 8:12 am
If you buy brokered CDs or treasuries, you can also sell at the click of a button.
Yes, but the cost of selling CDs is pretty high, maybe around 1% of the value.

I am using brokered CDs in IRA as a substitute for treasuries, so mixing those with non-index bond funds that have minimal treasuries.

We don't need CDs to be liquid. Can get that from our bond and other funds, as well as some >5 year old I-bonds.
press on, regardless - John C. Bogle

aristotelian
Posts: 4767
Joined: Wed Jan 11, 2017 8:05 pm

Re: CDs role in Fixed income part of portfolio

Post by aristotelian » Sat Sep 15, 2018 9:46 am

munemaker wrote:
Sat Sep 15, 2018 8:12 am
If you buy brokered CDs or treasuries, you can also sell at the click of a button.
I'll be honest, I have never done it. Can you really a) find a buyer instantaneously and b) know you are getting market price? Those would be my concerns. Like I said, I am becoming a convert to the CD ladder, but just for convenience I am keeping some bond funds at the ready for rebalancing and convenience.

keepingitsimple
Posts: 143
Joined: Sun Mar 18, 2018 8:20 pm

Re: CDs role in Fixed income part of portfolio

Post by keepingitsimple » Sat Sep 15, 2018 10:22 am

venkman wrote:
Fri Sep 14, 2018 9:05 pm
Right now, you can build a 5-year CD ladder with brokered CD's at Vanguard that yields 3.01% in the first year. If you could reinvest the maturing 1-year CD at current 5-year rates, the average yield goes up to 3.19% in year two. The average maturity of the ladder will be 2.5 years, and your principal is effectively as safe as it would be in Treasuries.

For comparison, VG's Short Term Treasury Index (VSBSX) has an average maturity of 2.0 years, and a current SEC yield of 2.58%.

If you have a significant amount in FI and don't need it all to be completely liquid, a CD ladder is a compelling possibility for at least part of your FI allocation, especially if the money isn't subject to state taxes. I'd still keep some assets in bond funds, just for easy rebalancing.
Additionally, to improve accessibility of funds, one could structure a 3-year or 5-year ladder with CDs maturing every six months.

JackoC
Posts: 261
Joined: Sun Aug 12, 2018 11:14 am

Re: CDs role in Fixed income part of portfolio

Post by JackoC » Sat Sep 15, 2018 10:22 am

aristotelian wrote:
Fri Sep 14, 2018 6:47 pm
Cyclesafe wrote:
Fri Sep 14, 2018 6:18 pm
CD's make sense for a retail investor who wants zero risk and accepts the lower return thus associated with them. Interim values of the CD's between purchase and maturity do fluctuate, however, just like bond funds. Sometimes buying CD's are better than buying similarly risk free treasury bonds (via Treasury Direct), and sometimes not - even when considering that states do not tax interest earned from treasuries.

The other options for fixed you mention are not risk free so they offer a premium return to compensate for this.

There's nothing whatsoever wrong with CD's being the sole investment in fixed. Many Bogleheads do this.
CDs do not necessarily have lower returns vs similar risk instruments. 5 year CDs typically beat Treasuries of the same duration. I would frame the choice as liquidity of Treasuries vs higher yield of CDs.
I'd go further. The *best* online direct CD yields in recent years have uniformly been much higher than comparable treasuries relatively speaking. It hasn't been implausible to average 100 bps in 5 yrs using every-once-in-a-while outlier best CD deals, little <60bps right now, especially narrow by recent standards. That overwhelms the state v federal taxation issue in almost all cases. It's a spread equivalent to taking serious credit or pre-payment risk in high grade bonds other than treasuries but you need take neither. Moreover the Early Withdrawal Penalty on CD's usually amounts to a put option of significant value in bps you don't get with treasuries. If rates skyrocket, you're just out of luck with treasuries (worse with the imbedded call options in a lot of non-treasuries in TBM funds) but the CD break clause could save significant money. The only way to get a worse yield on a CD than comparable treasury is to go down the list of CD's at a site like Depositaccounts and take significantly worse than the best one, or some even crummier CD rate at a local bank. But why? If you don't care about yield put it all in a treasury bond fund not a crummy rate CD, I'd agree.

You're getting paid the 60-100 bps (plus put option) for less liquidity basically plus a little work setting up, shifting around and keeping track of accounts, besides a debatable but in any case small credit risk difference*. It's not money for nothing, but a fat liquidity premium. You arrange the CD portion of fixed income to be that which doesn't require instant liquidity. If you're 60/40 bonds and the stock market declines 70% you still only need to shift 17 or so of the original 40 from bonds to stocks to rebalance to 60/40. And would you really keep rebalancing full blast if the market went down that far? I would not, don't plan to even a priori, I'll let somebody else be the hero in that case :D . So there's no reason in general for all your fixed income to be instantly liquid, though there are always exceptions due to special personal circumstances. And none of this considers whether high grade municipal bonds give a higher enough TE yield to offset their greater credit risk, which in high tax brackets they might depending also on credit risk preference.

*could arguably be in favor of CD's in some scenarios because of the dual credit nature. If the govt someday puts a 'haircut' on principal repayments of treasuries as a part of a solution to a fiscal crisis, a measure viewed by then as less bad than ramping up inflation on purpose to lower the real value of the principal, that's not necessarily going to apply to CD's, nor would the bond haircut make every bank in the country go bust. In fact there could be a corresponding bailout so 'small' depositors in banks don't get hurt, but 'fat cats' mainly bear the burden.

User avatar
vineviz
Posts: 2005
Joined: Tue May 15, 2018 1:55 pm

Re: CDs role in Fixed income part of portfolio

Post by vineviz » Sat Sep 15, 2018 11:06 am

michaeljc70 wrote:
Fri Sep 14, 2018 6:03 pm
I'm interested in what role CDs should play in a FI part of a portfolio, particularly near/in retirement compared to Total Bond or similar. Obviously, CDs don't lose principal value while they also don't gain outsized returns in volatile/dropping rate environments like longer term bonds (like during and after 2008). How do CDs compare long term to the performance of Total Bond (obviously duration of the CD comes into play)? Do you allocate a % of FI to CDs or X months of living expenses or????

I'd rather not fool with CDs/CD ladders, but am questioning that. My AA is 75/25 and I don't plan on changing that so maybe it is moot with only 25% FI.
Historically, the best overall INVESTMENT portfolio performance (i.e. highest risk-adjusted return) has come from holding intermediate or long-term bonds rather than short-term bonds, CDs, or cash in the fixed-income portion of your portfolio.

Obviously your short-term spending needs should be in savings vehicles (like CDs or money markets) rather than investment vehicles (like stocks and bonds), but if we are talking about the actual investment portfolio then replacing Total Bond with CDs will probably increase the risk and lower the return of the overall portfolio.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

JackoC
Posts: 261
Joined: Sun Aug 12, 2018 11:14 am

Re: CDs role in Fixed income part of portfolio

Post by JackoC » Sat Sep 15, 2018 11:50 am

vineviz wrote:
Sat Sep 15, 2018 11:06 am
Historically, the best overall INVESTMENT portfolio performance (i.e. highest risk-adjusted return) has come from holding intermediate or long-term bonds rather than... CDs...
I'm wondering what source could have concluded that and how it could be valid, especially on a forward looking basis, even if it did. As just mentioned, you can say 5yr CD's 'aren't necessarily lower yielding than the 5 yr note' if you take the *national average* of CD rates. If you take the best CD rates, the CD yield is significantly higher. Taking the recent situation where for example my CD portfolio was transacted at almost 100bps over treasuries on average. I wonder how a past study could account for that, where to get the data on what the *best* CD rates were far back in time. And maybe the spread wasn't as high in the distant past, but that's not directly relevant to now and the recent past when it has been.

Data problem number 2. The last 30yr plus years have seen a huge drop in yields, making high duration seem a lot more attractive than if you truncated the last 30+ years. Even a 5 yr *treasury* v 30 yr treasury comparison is subject to that recency bias. From where rates are now, it's pretty much impossible to replicate that huge bond bull market, though nobody knows if today's low rates are a permanent new normal or if we see much higher rates again, in which case the 30 yr will not be the place to be.

Data problem 3 is a source which includes the value of early withdrawal options on CD's historically. I'm pretty sure that doesn't exist. But it's a significant add on to the 60-100bp advantage of best 5 yr CD's over the 5 yr note. How does a 3.13% 30 yr (where you get slaughtered if rates jump) beat a 4% 5yr CD with break option (I bought one only months ago), in risk/return? That seems unlikely. Again, looking at past history using some estimate of average CD rates much closer to or even below the treasury curve, and ignoring the break option, isn't very relevant to deciding how likely.

Problem 4 is related to 2, another recency bias. In long term history long term treasury and stock prices have been essentially uncorrelated, like CD's (price=100 all the time) and stock prices. Only fairly recently has there been significant negative correlation and it could just as easily go away or turn positive, like it was in the period of unexpected inflation in the US in 1960's-early 80's: long term bonds and the stock market would both get hit by bad inflation news.

I don't think there's valid historical evidence to conclude that the best CD rates now would lower expected return or increase risk relative to piling on treasury duration risk. And it's as straight forward as can be to see that holding the 5 yr note to maturity has less return than the best 5 yr CD's with basically the same credit risk. Again as long as one uses CD's in a portion of their FI portfolio which does not need instant liquidity.

michaeljc70
Posts: 3746
Joined: Thu Oct 15, 2015 3:53 pm

Re: CDs role in Fixed income part of portfolio

Post by michaeljc70 » Sat Sep 15, 2018 12:05 pm

JackoC wrote:
Sat Sep 15, 2018 11:50 am
vineviz wrote:
Sat Sep 15, 2018 11:06 am
Historically, the best overall INVESTMENT portfolio performance (i.e. highest risk-adjusted return) has come from holding intermediate or long-term bonds rather than... CDs...
I'm wondering what source could have concluded that and how it could be valid, especially on a forward looking basis, even if it did. As just mentioned, you can say 5yr CD's 'aren't necessarily lower yielding than the 5 yr note' if you take the *national average* of CD rates. If you take the best CD rates, the CD yield is significantly higher. Taking the recent situation where for example my CD portfolio was transacted at almost 100bps over treasuries on average. I wonder how a past study could account for that, where to get the data on what the *best* CD rates were far back in time. And maybe the spread wasn't as high in the distant past, but that's not directly relevant to now and the recent past when it has been.

Data problem number 2. The last 30yr plus years have seen a huge drop in yields, making high duration seem a lot more attractive than if you truncated the last 30+ years. Even a 5 yr *treasury* v 30 yr treasury comparison is subject to that recency bias. From where rates are now, it's pretty much impossible to replicate that huge bond bull market, though nobody knows if today's low rates are a permanent new normal or if we see much higher rates again, in which case the 30 yr will not be the place to be.

Data problem 3 is a source which includes the value of early withdrawal options on CD's historically. I'm pretty sure that doesn't exist. But it's a significant add on to the 60-100bp advantage of best 5 yr CD's over the 5 yr note. How does a 3.13% 30 yr (where you get slaughtered if rates jump) beat a 4% 5yr CD with break option (I bought one only months ago), in risk/return? That seems unlikely. Again, looking at past history using some estimate of average CD rates much closer to or even below the treasury curve, and ignoring the break option, isn't very relevant to deciding how likely.

Problem 4 is related to 2, another recency bias. In long term history long term treasury and stock prices have been essentially uncorrelated, like CD's (price=100 all the time) and stock prices. Only fairly recently has there been significant negative correlation and it could just as easily go away or turn positive, like it was in the period of unexpected inflation in the US in 1960's-early 80's: long term bonds and the stock market would both get hit by bad inflation news.

I don't think there's valid historical evidence to conclude that the best CD rates now would lower expected return or increase risk relative to piling on treasury duration risk. And it's as straight forward as can be to see that holding the 5 yr note to maturity has less return than the best 5 yr CD's with basically the same credit risk. Again as long as one uses CD's in a portion of their FI portfolio which does not need instant liquidity.
Though I don't have historical data on CDs going back really far, when looking at the 10 year treasury I can see that there appears to be only one year (since 1928) where the 10 year lost double digits while there are 18 years where it had double digit returns. Many were also in low inflation years. I don't know if these gains (along with the losses) amount to more than a 5 year CD over long periods. I would guess they do.

User avatar
vineviz
Posts: 2005
Joined: Tue May 15, 2018 1:55 pm

Re: CDs role in Fixed income part of portfolio

Post by vineviz » Sat Sep 15, 2018 1:33 pm

JackoC wrote:
Sat Sep 15, 2018 11:50 am
I'm wondering what source could have concluded that and how it could be valid, especially on a forward looking basis, even if it did. As just mentioned, you can say 5yr CD's 'aren't necessarily lower yielding than the 5 yr note' if you take the *national average* of CD rates. If you take the best CD rates, the CD yield is significantly higher.
Obviously it's impossible for the average investor, much less all investors, to get the highest available CD rates.

Furthermore, the last data I saw noted that the average fixed rate CD is issued with a maturity of less than 12 months and the average variable rate CD is issued with a maturity of less than 24 months. As a result, the average CD portfolio has a MUCH shorter average maturity than the average bond portfolio. A rolling ladder of 5-year CDs would have an average effective maturity of 2.5 years whereas the Vanguard Total Bond Market fund has an average effective maturity of 8.4 years
JackoC wrote:
Sat Sep 15, 2018 11:50 am
The last 30yr plus years have seen a huge drop in yields, making high duration seem a lot more attractive than if you truncated the last 30+ years.
This is a popular misconception. Long duration portfolios tend to have higher total returns than short duration portfolios, even when interest rates are rising. https://www.pimco.com/en-us/insights/vi ... e-long-run

And in any intersest rate environments, long-duration bonds are stronger diversifiers for stocks than short-duration bonds
JackoC wrote:
Sat Sep 15, 2018 11:50 am
Data problem 3 is a source which includes the value of early withdrawal options on CD's historically. I'm pretty sure that doesn't exist.
I'd like to see that data, but as a first order approximation I think it's naive to assume that banks are persistently misplacing that optionality.
JackoC wrote:
Sat Sep 15, 2018 11:50 am
Only fairly recently has there been significant negative correlation and it could just as easily go away or turn positive, like it was in the period of unexpected inflation in the US in 1960's-early 80's: long term bonds and the stock market would both get hit by bad inflation news.
Predicting the future is difficult, so we really have no way of knowing what would happen if we see a bout of high inflation again. Nonetheless, the Fed under Volcker made some pretty important changes in the early 1980s in their targets, shifting from "stop and go" interest rate targets to inflation targets. We could experience a reversal in monetary policy again, I suppose, but unless we do the evidence I've seen suggests that it is reasonable to expect the duration premium to persist. We've have experience several periods of fed funds rate increases since 1980, and long-term bonds outperformed in subsequent periods far more often than they underperformed.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

venkman
Posts: 708
Joined: Tue Mar 14, 2017 10:33 pm

Re: CDs role in Fixed income part of portfolio

Post by venkman » Sat Sep 15, 2018 9:54 pm

munemaker wrote:
Sat Sep 15, 2018 8:15 am
venkman wrote:
Fri Sep 14, 2018 9:05 pm
VG's Short Term Treasury Index (VSBSX) has an average maturity of 2.0 years, and a current SEC yield of 2.58%.
Note: The YTD return of VSBSX is 0.10%. The one year total return is -0.31%.

A CD ladder purchased a year ago would likely have a similar total return. SEC yield is a far more relevant predictive number for bonds than past return.

NoblesvilleIN
Posts: 19
Joined: Sat Mar 24, 2018 6:04 pm
Location: Noblesville IN

Re: CDs role in Fixed income part of portfolio

Post by NoblesvilleIN » Sun Sep 16, 2018 9:44 am

I allocate a % of living expense to CD's. I have a CD ladder of 8 CD's with one maturing in the 3rd month of each quarter (2 year ladder). Each CD's $ value is 1/4th of a year's expense. My thought is that in a downturn where my other investments don't produce the expected income, I can expect a CD to mature within the next 2 or 3 months and get me by. If I don't need it, I roll it into another 2 year CD. My CD's are kept in my taxable brokerage account (Fidelity) and I use their CD tool to find the best rate. So, once a quarter I have to spend a little time looking at CD rates.

I figure each quarter I have 4 options: 1) Keep the money (which means in 2 years, I won't have that option for a particular quarter); 2) Roll the CD for another 2 years; 3) Keep some of the money and roll the rest (which means in 2 years, I'll have less to roll or spend); 4) Add some savings to the CD when I roll (which means in 2 years, I'll have more to spend or roll).

I am currently doing option 4 - adding to each CD some savings. I plan to retire at the end of the year and will decide next year whether to keep adding to the CD's or not.

The reason I picked the 3rd month of a quarter for my CD purchases is because I have noticed over the last 5 years of tracking investment income, that the income is lumpy from month to month, but smoother over a quarter. My thought is that what I collect from this quarter's investment income is what will be available to spend next quarter (divided by 3 for monthly spend). By the 3rd month of the quarter, I should have a good idea if I'm going to need funds from a CD or not and if I have "extra" to add to a CD.

I'm hoping this will allow me to ride out a rough patch for a couple of years before I have to worry about selling something (bonds or stocks). I also keep about a quarter's worth of income in a money market at the brokerage as a float.

UpperNwGuy
Posts: 1110
Joined: Sun Oct 08, 2017 7:16 pm

Re: CDs role in Fixed income part of portfolio

Post by UpperNwGuy » Sun Sep 16, 2018 9:49 am

CDs have no role in my portfolio. I use Vanguard money market funds for short-term savings, and bond funds for everything else.

billthecat
Posts: 284
Joined: Tue Jan 24, 2017 2:50 pm

Re: CDs role in Fixed income part of portfolio

Post by billthecat » Sun Sep 16, 2018 10:29 am

UpperNwGuy wrote:
Sun Sep 16, 2018 9:49 am
CDs have no role in my portfolio. I use Vanguard money market funds for short-term savings, and bond funds for everything else.
That’s what I was thinking. MM for the cash portion of my total assets (5%), total bond market for the investment portion of fixed income (35%).

I may max out the annual I-bonds just because the yield can be deferred, though. Better for taxes. It’s no a lot though, at only 10-15k per year (which I would deduct from the 35% fixed income).

User avatar
friar1610
Posts: 1290
Joined: Sat Nov 29, 2008 9:52 pm
Location: MA South Shore

Re: CDs role in Fixed income part of portfolio

Post by friar1610 » Sun Sep 16, 2018 11:19 am

NoblesvilleIN wrote:
Sun Sep 16, 2018 9:44 am
I allocate a % of living expense to CD's. I have a CD ladder of 8 CD's with one maturing in the 3rd month of each quarter (2 year ladder). Each CD's $ value is 1/4th of a year's expense. My thought is that in a downturn where my other investments don't produce the expected income, I can expect a CD to mature within the next 2 or 3 months and get me by. If I don't need it, I roll it into another 2 year CD. My CD's are kept in my taxable brokerage account (Fidelity) and I use their CD tool to find the best rate. So, once a quarter I have to spend a little time looking at CD rates.

I figure each quarter I have 4 options: 1) Keep the money (which means in 2 years, I won't have that option for a particular quarter); 2) Roll the CD for another 2 years; 3) Keep some of the money and roll the rest (which means in 2 years, I'll have less to roll or spend); 4) Add some savings to the CD when I roll (which means in 2 years, I'll have more to spend or roll).

I am currently doing option 4 - adding to each CD some savings. I plan to retire at the end of the year and will decide next year whether to keep adding to the CD's or not.

The reason I picked the 3rd month of a quarter for my CD purchases is because I have noticed over the last 5 years of tracking investment income, that the income is lumpy from month to month, but smoother over a quarter. My thought is that what I collect from this quarter's investment income is what will be available to spend next quarter (divided by 3 for monthly spend). By the 3rd month of the quarter, I should have a good idea if I'm going to need funds from a CD or not and if I have "extra" to add to a CD.

I'm hoping this will allow me to ride out a rough patch for a couple of years before I have to worry about selling something (bonds or stocks). I also keep about a quarter's worth of income in a money market at the brokerage as a float.
This is very close to a strategy I have been mulling over except I would do it within an IRA and the CD amounts - probably in a 3 year ladder - would be pegged to anticipated RMDs. The rest of the IRA would remain in a Total Bond Mkt fund. I know some recommend either Prime MMF or short-term bond funds for this "safest" portion of a safe IRA portfolio and that's what I'm still trying to figure out. The options you described are similar to what I came up with except that, being of RMD age, I don't have an option just to leave the money in the IRA.
Friar1610

JackoC
Posts: 261
Joined: Sun Aug 12, 2018 11:14 am

Re: CDs role in Fixed income part of portfolio

Post by JackoC » Sun Sep 16, 2018 12:05 pm

vineviz wrote:
Sat Sep 15, 2018 1:33 pm
JackoC wrote:
Sat Sep 15, 2018 11:50 am
I'm wondering what source could have concluded that and how it could be valid, especially on a forward looking basis, even if it did. As just mentioned, you can say 5yr CD's 'aren't necessarily lower yielding than the 5 yr note' if you take the *national average* of CD rates. If you take the best CD rates, the CD yield is significantly higher.
1. Obviously it's impossible for the average investor, much less all investors, to get the highest available CD rates.

2. Furthermore, the last data I saw noted that the average fixed rate CD is issued with a maturity of less than 12 months and the average variable rate CD is issued with a maturity of less than 24 months. As a result, the average CD portfolio has a MUCH shorter average maturity than the average bond portfolio. A rolling ladder of 5-year CDs would have an average effective maturity of 2.5 years whereas the Vanguard Total Bond Market fund has an average effective maturity of 8.4 years
JackoC wrote:
Sat Sep 15, 2018 11:50 am
The last 30yr plus years have seen a huge drop in yields, making high duration seem a lot more attractive than if you truncated the last 30+ years.
3. This is a popular misconception. Long duration portfolios tend to have higher total returns than short duration portfolios, even when interest rates are rising. https://www.pimco.com/en-us/insights/vi ... e-long-run

And in any intersest rate environments, long-duration bonds are stronger diversifiers for stocks than short-duration bonds
JackoC wrote:
Sat Sep 15, 2018 11:50 am
Data problem 3 is a source which includes the value of early withdrawal options on CD's historically. I'm pretty sure that doesn't exist.
4. I'd like to see that data, but as a first order approximation I think it's naive to assume that banks are persistently misplacing that optionality.
JackoC wrote:
Sat Sep 15, 2018 11:50 am
Only fairly recently has there been significant negative correlation and it could just as easily go away or turn positive, like it was in the period of unexpected inflation in the US in 1960's-early 80's: long term bonds and the stock market would both get hit by bad inflation news.
5. Predicting the future is difficult, so we really have no way of knowing what would happen if we see a bout of high inflation again. Nonetheless, the Fed under Volcker made some pretty important changes in the early 1980s in their targets, shifting from "stop and go" interest rate targets to inflation targets. We could experience a reversal in monetary policy again, I suppose, but unless we do the evidence I've seen suggests that it is reasonable to expect the duration premium to persist. We've have experience several periods of fed funds rate increases since 1980, and long-term bonds outperformed in subsequent periods far more often than they underperformed.
Several points but my response comes down to two. For your points 1, 2, and 4 you are making an assumption which is often reasonable for individuals dealing in an efficient market but is not in any of these three cases. For example, it's reasonable to say that the general assumption 'investors can't beat the stock market by timing' is applicable to any given individual. Not 100%, but in general. But the CD market can't be arbitraged by professionals so this assumption is not valid.

Anyone can go to depositaccounts.com, look for exceptional deals to pop up and take them. Again if this were the professional market, everyone would do that, and the difference between good and bad CD deals would be arbed to almost nothing. In the real world the great majority of CD investor don't pay attention to those sites or the yield at all with a broad range. So an individual investor who does care can get much better deals. A realistic analysis based on 'individually actionable' has to take that into account.

Similarly and more obviously with the term of CD's. It's up to the individual to choose. If in fact very short maturities are less efficient, it's irrelevant, and arbitrary to compare based on an average.

Finally with the break options. There is no reason to believe those are efficiently priced because again they are given to retail investors who don't exercise them optimally, often breaking when rates have actually gone down, because they need the money. It's the same reason the mortgage prepayment option is manifestly cheap to a rational borrower exercising strictly according to interest rates, because few do, and even rational borrowers often can't (have to move to get a job, etc). Even call options embedded in public bonds are traditionally cheap...because so many of them go to retail including as part of TBM funds and retail investors tend to ignore options they are short as well as non-optimally exercising those they are long. It is not valid to assume you 'get what you pay for' on CD break options. In fact it's obviously not true. The yield is up 100bp superior to treasuries for basically the same credit risk *before* considering the value of the break option, which could be as much as another 20bp or more.

So your (it seems hypothetical actually) source that says bonds are a superior investment to CD's would in fact have major data problems in dealing with how individuals could get significantly higher than treasury yields on CD's, and benefit from break options. And it would arbitrary for this source, if it existed, to choose short maturity CD's.

On points 3 and 5 your link and arguments seem to depend in part on the flawed assumption in 2: as already mentioned the investor has no obligation to choose CD maturities any less than 5 yrs, and if term premium is a big factor beyond that can in fact find CD's longer, still well above the treasury curve. The linked paper is also not a historical analysis but a simulation starting from the very steep curve of a few years ago. Today's curve is comparatively flat and the term premium is by some estimates actually negative (note: term premium on ex-ante basis meaning the difference between true expected future spot rate and today's forward rate between given stationary calendar dates, term premium on realized basis meaning the difference between the realized future spot rate and current forward between given dates). Anyway the investor is also free to choose different fixed income investing methods as conditions vary.

When I say loads of duration has done better in the long bond bull market I mean the 30 yr. And the big bond bull market of the last few decades *would* be a significant factor in historical analysis (which is what you originally suggested 'showed' bonds superior to CD's) that called 'bonds' the 30 yr.

Anyway for fairly similar maturities it's pretty obvious: a puttable CD 100+ over the curve in ~5yrs (such as I bought a few months ago) is not going to have a lower expected return than 8-10 treasury flat, held constant maturity (rolling) or to maturity. The yield spread and option advantage is just too big. Other cases might be closer run. An all in fair historical analysis would be very difficult to construct, and I doubt even more than before that one exists.

JackoC
Posts: 261
Joined: Sun Aug 12, 2018 11:14 am

Re: CDs role in Fixed income part of portfolio

Post by JackoC » Sun Sep 16, 2018 12:18 pm

michaeljc70 wrote:
Sat Sep 15, 2018 12:05 pm


Though I don't have historical data on CDs going back really far, when looking at the 10 year treasury I can see that there appears to be only one year (since 1928) where the 10 year lost double digits while there are 18 years where it had double digit returns. Many were also in low inflation years. I don't know if these gains (along with the losses) amount to more than a 5 year CD over long periods. I would guess they do.
I would not necessarily guess so if the CD had a yield over the treasury curve like the 60-100bps that's prevailed on best deals in recent years, and took into account breaking and replacing CD's when rates spiked. Return over time that is. Mark to market gain of mark to market instrument v non mark to market instrument over a short period is apples and oranges. If the 10 yr rate drops from 3% to 2% you record a 9% (give or take) MTM gain on a treasury but thereafter get 2% yield. The CD you bought at 4% has no MTM gain but now accrues 2% faster than the treasury rather than 1%. So if the idea is double digit MTM gains 10yr notes have given that 'a CD could never give you', that reasoning is off I believe. But anyway if the comparison was under the assumption 5yr CD=5yr treasury yield, no optionality, v the 10 yr treasury I also would guess 10 yr would win historically on a correct counting basis. But to summarize and simplify the previous exchange, if you ruled out important yield and optionality differences between best CD deals (any particular investor can get even if most don't care enough) and treasuries... but that wouldn't be a useful analysis on a forward looking individual basis.

michaeljc70
Posts: 3746
Joined: Thu Oct 15, 2015 3:53 pm

Re: CDs role in Fixed income part of portfolio

Post by michaeljc70 » Sun Sep 16, 2018 12:43 pm

JackoC wrote:
Sun Sep 16, 2018 12:18 pm
michaeljc70 wrote:
Sat Sep 15, 2018 12:05 pm


Though I don't have historical data on CDs going back really far, when looking at the 10 year treasury I can see that there appears to be only one year (since 1928) where the 10 year lost double digits while there are 18 years where it had double digit returns. Many were also in low inflation years. I don't know if these gains (along with the losses) amount to more than a 5 year CD over long periods. I would guess they do.
I would not necessarily guess so if the CD had a yield over the treasury curve like the 60-100bps that's prevailed on best deals in recent years, and took into account breaking and replacing CD's when rates spiked. Return over time that is. Mark to market gain of mark to market instrument v non market to market instrument over a short period is apples and oranges. If the 10 yr rate drops from 3% to 2% you record a 9% (give or take) MTM gain on a treasury but thereafter get 2% yield. The CD you bought at 4% has no MTM gain but now accrues 2% faster than the treasury rather than 1%. So if the idea is double digit MTM gains 10yr notes have given that 'a CD could never give you', that reasoning is off I believe. But anyway if the comparison was under the assumption 5yr CD=5yr treasury yield, no optionality, v the 10 yr treasury I also would guess 10 yr would win historically on a correct counting basis. But to summarize and simplify the previous exchange, if you ruled out important yield and optionality differences between best CD deals (any particular investor can get even if most don't care enough) and treasuries... but that wouldn't be a useful analysis on a forward looking individual basis.
If you are going to break CDs at some point, the analysis becomes even more difficult. Obviously everyone isn't getting the best rate on a CD. Most people like to keep it local rather than sending their money to some bank in another state they never heard of.

One thing I will say is a reason that treasuries had such good returns during the crisis was foreign investors. They cannot easily buy CDs in the US. I didn't hear of people flocking to CDs. There is no maximum number of CDs while treasuries are limited (though I know the amount out there is huge).

JackoC
Posts: 261
Joined: Sun Aug 12, 2018 11:14 am

Re: CDs role in Fixed income part of portfolio

Post by JackoC » Sun Sep 16, 2018 1:35 pm

michaeljc70 wrote:
Sun Sep 16, 2018 12:43 pm

1. If you are going to break CDs at some point, the analysis becomes even more difficult. Obviously everyone isn't getting the best rate on a CD. Most people like to keep it local rather than sending their money to some bank in another state they never heard of.

2. One thing I will say is a reason that treasuries had such good returns during the crisis was foreign investors. They cannot easily buy CDs in the US. I didn't hear of people flocking to CDs. There is no maximum number of CDs while treasuries are limited (though I know the amount out there is huge).
1. Yes a difficult analysis, but that doesn't make it valid to talk about 'past history says bonds do better' without a complete analysis. Again, the 'but everybody isn't going to get the best deal' argument is fundamentally less valid in context of an inefficient retail market than in the context of efficient professional market from which it comes. 'Some investors will beat the stock market by timing but on average they can't' is because to beat that market you have to beat professional smart money and it's reasonable to assume that's, at best, a random chance. To beat the average CD investor on rate all you have to do is pull up depositaccounts and select the best deal, with a little homework that tells you the institution is just as legit as the local one. A little further homework and diligence gets you a portfolio of CD's with higher spreads than the best ones on DA at any given time on average. This is simply not the same as claiming you can beat the stock market, and it's just basically irrelevant in this case to say 'well everyone doesn't do that'. So what? From an 'individually actionable' POV.

2. Again I think apples v. oranges of MTM v non-MTM instrument must enter in here. If you buy a CD at 70 over comparable treasuries, then foreigners rush into the treasury market and push up prices, you are still 70bps ahead, the price rise in the treasury is exactly offset by lower yield from here on, a tautology. If OTOH you want to buy treasuries today but see that foreigners rushed in yesterday and pushed yields down 10bps, yesterday's best CD offer may very well still be there and now you can buy the CD 80 over treasuries. In the professional market banks would get immolated doing that, but in the price insensitive inefficient retail CD market they often leave up the same rates when the bond market has rallied (or sold off) significantly in recent days: most retail investors aren't paying enough attention take advantage of the bank when market rates suddenly drop, they are just as likely to take yesterday's best CD rate when bond yields went up 20bps recently rather than wait for CD's to reflect it. But moving rates every day would annoy them and seem dishonest 'bate and switch'. I've often bought CD's just after big treasury rallies, before they go down.

User avatar
Johnnie
Posts: 512
Joined: Sat May 28, 2016 3:18 pm
Location: Michigan

Re: CDs role in Fixed income part of portfolio

Post by Johnnie » Sun Sep 16, 2018 6:36 pm

JackoC wrote:
Sat Sep 15, 2018 11:50 am
I'm wondering what source could have concluded that and how it could be valid, especially on a forward looking basis, even if it did. As just mentioned, you can say 5yr CD's 'aren't necessarily lower yielding than the 5 yr note' if you take the *national average* of CD rates. If you take the best CD rates, the CD yield is significantly higher...
JackoC, the points in this and your other posts here are exactly what has been going through my head the last year or so, as I'm getting ever closer to retirement and starting to back my A/A down from 60/40 to 50/50. What you say seems intuitive and common-sensical to me, while the counter arguments often seem counter-intuitive and abstract.

I was previously unaware that some of these higher-yielding CDs allow investors to bail early with no penalty. I can't picture needing to do that myself, but it seems to wipe out the biggest drawback to CDs, the loss of liquidity.

I am also concerned about the chances that the bond owner's multi-decade summer of declining rates may give way to a long winter of rising rates - a winter that at my age might as well be forever. Like a lot of people I was "early is wrong" on that belief, and per my signature line I don't put too much weight on this opinion, but the "what's the most you can lose?" question suggests I would rather be wrong by keeping duration shorter than wrong by going longer. Higher CD yields make that choice easier too.

One other factor is something Larry Swedroe said recently, that if the stuff hits the fan a' la' 2000 and 2008 the "investment grade" bond funds will take a bit hit and the 100% treasury ones will get a bounce. Given that bonds are in my portfolio for safety, that has me inclined to move in the Treasury direction, at the cost of slightly lower yields. CDs would not provide a similar bounce in a big down-draft, but the fact that they are just as safe as Treasuries and come with higher yields means I don't care very much about that.

I stuck my toe in a few months ago buying highest-yield CDs in an IRA at Schwab and will probably do more.

Thanks
"I know nothing."

krafty81
Posts: 75
Joined: Thu Jul 05, 2018 2:01 pm
Location: San Diego, CA

Re: CDs role in Fixed income part of portfolio

Post by krafty81 » Sun Sep 16, 2018 9:41 pm

I am beginning to purchase CDs again. Someone recommended Andrews FCU and they have some amazing rates. I learned after investing that you can only buy one per purchaser, kind of wish I had invested more. If I know I will not need the money, I just don't want to lose to inflation and bonds are not impressing me right now.

Nodrog
Posts: 6
Joined: Tue Jun 06, 2017 4:27 am
Location: Australia

Re: CDs role in Fixed income part of portfolio

Post by Nodrog » Mon Sep 17, 2018 12:14 am

A combination of CDs and Bond Fund?:

https://www.moneysense.ca/save/retireme ... nt-income/

As a retiree I personally only use CDs (Term Deposits in Australia) for the risk free allocation of the portfolio.

User avatar
Top99%
Posts: 332
Joined: Sat Apr 22, 2017 9:30 am
Location: Austin, TX

Re: CDs role in Fixed income part of portfolio

Post by Top99% » Mon Sep 17, 2018 7:41 am

I am also a fan of CD ladders as part of ones fixed income portfolio. I just checked how our rolling 1-2-3 year CD ladder we setup during 2005 for our emergency fund did and it has outpaced inflation by a little bit. So, meeting our expectations.
Adapt or perish

Post Reply