Too Good to Be True?

[stuper1]

You know the old saying from a business person to a customer:  "good, fast, or cheap -- pick two". 

From what I'm reading about the PP, it seems like you get all three.  You get good performance with much lower fluctuations than most portfolios.  The PP is fast in terms of not requiring much personal time at all for upkeep.  And it's cheap in terms of a very low expense ratio.  I'm almost 100% invested in the PP now, and my overall expense ratio looks like it will be about 0.15%, including some miscellaneous costs such as ETF commissions and a safe deposit box for gold.  I'm sure others have overall expense ratios even better; mine suffers a bit from having most of my money in a 401k and IRA.

It seems too good to be true.  I know there is no guarantee of future results, but this sounds a lot more likely to produce good returns with low time and expenses than a lot of other strategies I've heard of.

[MediumTex]

It only seems too good to be true because of all of the sloppy investment thinking that came before it in most cases (that was certainly true in my case).

A year from now you will have much more subtle insights into whether the PP is really too good to be true.

As long as you don't mind 100% stock investors occasionally running away from you in terms of returns, the PP can provide outstanding long term investment results.

[melveyr]

You know the saying more risk = more reward? Well that is not true under CAPM or any decent model of asset returns. The correct statement is more compensated risk = more expected reward.

Most investors take uncompensated risk, risks that the capital market has never rewarded. Owning a single stock is an extreme example. That risk is diversifiable so the capital markets do not reward the business specific risks. Most Bogleheads understand this, and own the total stock market to diversify away from a lot of the uncompensated risk.

However, even owning the total stock market is taking on uncompensated risk. Bonds help diversify against many of the equity risks, most notable deflation. A stock/bond portfolio leveraged up to the risk level of the stock market has higher returns. A 100% equity portfolio is just dumb.

Now, the stock/bond portfolio is vulnerable to inflation. Should investors be rewarded for bearing that risk? Not entirely, because gold is a pretty decent diversifier for this scenario. I think that even a stock/bond portfolio is taking on some uncompensated risk.

I guess the summary is that lots of people suck at investing and make portfolios that don't make sense and therefore the capital markets do not reward these investors. The PP is not perfect but it gets closer than lots of other portfolios and we are lucky enough to live in a time where frictional trading costs are mere rounding errors.

[D1984]

The Permanent Portfolio isn't too good to be true...IF:

1. You don't mind at some point in your investing career earning maybe 1 or 2% real (after inflation) per year for a decade or two on end. The PP has never really been fully tested in an environment where gold and bonds were both getting taken to the cleaners over a period of more than a year or two (the closest we have is what happened in the 1980-81 and 81-82 tight money recessions and that wasn't very pretty for the PP...the only thing that kept it from being worse was that interest rates were at 16% or more and stocks were already at generational lows in PE/10 terms so they didn't fall too much further) with only equities rising to offset them. The scenario would be a period of prosperity with positive (and overall rising) real rates and an equity bull market but the rising stock values wouldn't be enough to offset the falling gold and bond values....think what happened to gold in the 80 and 90s combined with what happened to LTTs in the 50s and 60s.

2. You don't mind actually losing money or earning 0% real on average for years in a row during a deflationary "lost decade" because apparently once LTT rates reach around 2% or thereabouts (see the Japanese PP for an example) they don't fall futher to offset--by giving a capital gain on the now higher than market rate bonds you possess--the effect of gold and stocks falling together (and cash doesn't really help much in such a situation because it earns little or nothing). LTT interest rates don't typically fall any further thanks to the fact that short term rates are already at the nominal zero lower bound and there will be at least some "interest rate risk premium" on holding LTTs vs STTs due to the fact that rates cannot go any lower but they most certainly can rise. If there was no such thing as physical currency then rates COULD go below zero and the zero lower bound might be a moot point but there is and it isn't.

3. You don't mind your three out of four assets losing value in a high interest rate tight money environment like the "Volcker shock" of the early 80s. STTs were already yielding double digits when that started....if we get another tight money shock from current rates it's going to be ugly for the PP.

[escafandro]

If we get another tight money shock from current rates it's going to be ugly for the PP.

Is there any way to measure this?

[D1984]

If we get another tight money shock from current rates it's going to be ugly for the PP.

Is there any way to measure this?

I apologize but I'm not quite sure what you're asking; by "any way to measure this" do you mean "is there any way to measure how bad it might be for the PP" or were you instead asking "is there any way to measure or determine when a tight money shock is a about to happen"?

[escafandro]

I apologize but I'm not quite sure what you're asking; by "any way to measure this" do you mean "is there any way to measure how bad it might be for the PP"...

This interpretation. 
Theoretically react of each component in isolation? And the portfolio as a whole?

[D1984]

I apologize but I'm not quite sure what you're asking; by "any way to measure this" do you mean "is there any way to measure how bad it might be for the PP"...

This interpretation. 
Theoretically react of each component in isolation? And the portfolio as a whole?

There's no way to measure it with perfect accuracy, no. One can make an educated guess, though:

First, we have to set our parameters. Let's say that rates on LTTs only go up to around 7% (which is where they roughly were as recently as 2000) and STTs rates get up to around the same level (in a tight money recession the yield curve flattens). Furthermore let's stipulate that inflation at the time is at maybe 3% so real interest rates are at maybe 4%. So it's not a truly tight money situation like we had in 1981 (10% inflation but 18% interest rates for an 8% real yield) but it's certainly a lot tighter than we currently have.

The potential results (all hypothetical guesses although the bond and SHY calcs were done based on ionterets rate formulas using bond calculators online):

Bonds (assuming 25 year average maturity and current rate of 3.14%) = a 45% capital loss which is slightly offset by the yield of 3.14% for a total return of -41.86%

Stocks = Anyone's guess; Bernanke's cheap money may have inflated stocks a little or a lot. If we assume dividend yields will have to go back to being maybe even 35% of 10-year bond yields (again, just a guess) and the 10-year is yielding perhaps 6.75% or 7% then the S&P 500 needs to yield around 2.43%. That would imply a loss of around 16% in the S&P 500. In 1981, the 10-year yield around 13.5% or 14% and the S&P yielded around 5.5 or 6%. If we use that ratio then it would imply that stocks should yield maybe 40% of what the 10-year pays which would mean a yield of around 2.75% which would mean a decline of around 25.8% in capital value for equities. Note that this is all assuming that almost no stocks cut their dividends despite the thight money recession. If they do the losses could be much worse. Let's just ssay a loss of 20% in capital value for stocks as a an average guess....20% minus 2.04% (the current yield which would offset the capital loss some) is roughly an 18% loss.

Gold = Who knows....the main factor is that gold HATES positive real rates much above 2%. IIRC gold lost around 35% in 1981 (it's worst year) and even when rates were far lower in the 80s and 90s there were years when it lost around 20%. Let's just say it guess it loses 26% in a tight money recession as described above.

Cash = No loss and a roughly 6.75% gain if you hold a savings account or a t-bill money market fund; roughly an 11.1% capital loss if you hald SHY (assuming a 1.75 year maturity and yields on it going from its current roughly 0.2% to 7.00%); the 11.1% capital loss would be offset by the increased yield so you would have a roughly 4.1% loss

So, for an ETF PP in a mild tight money recession the results might be as follows:

TLT = -41.86%
SPY = -18.00%
GLD = -26.00%
SHY = -4.10%

For an average loss of roughly -22.5% nominal or -25.5% real. That's pretty bad for a "stable, low-risk, conservative allocation" type of portfolio. Also, keep in mind that the "tight money" recession that I postulated was a rather mild one by historical terms...not as bad even as 1969-70 and certainly not as awful as 1980-81. If rates do rise quickly to double digits (say inflation gets up to maybe 5% and Bernanke's successor decides that austerity and tight money is just the medicine we need) then I don't even want to think about what that will do to an unmodified PP. The problem I see is that HB never really backtested the PP in a time when rates were as low as they are now; even if you start in 1968 or 1970 instead of 1972 rates were still at around 5.5% or 6% (or higher); a bond paying, say, 7% is going to be a lot less effected by an interest rate increase of, say, 200 basis points than one paying 3% is.

[escafandro]

So, for an ETF PP in a mild tight money recession the results might be as follows:

TLT = -41.86%
SPY = -18.00%
GLD = -26.00%
SHY = -4.10%

Scary picture.
So no assets would remain relatively afloat...
And the only thing one could try is to cut losses then?

[MachineGhost]

I guess the summary is that lots of people suck at investing and make portfolios that don't make sense and therefore the capital markets do not reward these investors. The PP is not perfect but it gets closer than lots of other portfolios and we are lucky enough to live in a time where frictional trading costs are mere rounding errors.

I'm wincing at your use of "model" and "reward" in the same sentence.  The only model that is working in the real world is the "rip the face off the muppets" model.  That's the real reward.  Just being in the market doesn't give you a reward, you must "rip the face off of a muppet" first even through the PP's volatility capturing.

[MachineGhost]

Cash = No loss and a roughly 6.75% gain if you hold a savings account or a t-bill money market fund; roughly an 11.1% capital loss if you hald SHY (assuming a 1.75 year maturity and yields on it going from its current roughly 0.2% to 7.00%); the 11.1% capital loss would be offset by the increased yield so you would have a roughly 4.1% loss

In 1981, a 5-year ladder or single 3-year of T-Notes returned 15.73%.  If weighted appropriately, it was enough to reduce the losses in the other three assets by half.  The FFR at the time was 20% while the CPI was at 13.5%, so there is a real return gain even if it looks like a loss in nominal terms.

[goodasgold]

How would I-bonds do in a tight money shock?

[sophie]

Nice analysis!

So, for an ETF PP in a mild tight money recession the results might be as follows:

TLT = -41.86%
SPY = -18.00%
GLD = -26.00%
SHY = -4.10%

For an average loss of roughly -22.5% nominal or -25.5% real.

Scary indeed, but two things to keep in mind:

- That's drawdown, not necessarily annual loss.  The PP has experienced such drawdowns before, but thankfully they're typically followed by a spike in returns.  After the 1 - 1.5 year painful episode, the market in at least one asset would rebound, and since you'll have rebalanced out of cash the portfolio overall will do well.  That in fact happened after 1981.

- What would a typical 50/50 Boglehead portfolio do in this situation?  Assuming total bond funds perform about halfway between TLT and SHY, there would be a very similar loss/drawdown, around 20%.  The only way to be completely protected from a loss would be to hold 100% cash. 

So I'm not dissuaded from the PP even in this nightmare scenario, although if long term bond rates get down to 1-1.5% I suspect there would be a lot of discussion about switching the bond allocation to cash or least intermediate term bonds.

[cowboyhat]

I think D1984 and Sophie make good points. Another to consider is that a tight-money recession is only one of the possible ways this may end. I think of it as the "responsible-adults-in-charge" ending, but there is also the "party-on-Wayne" ending where bond prices remain high for the rest of your time on earth. Seems unlikely, but consider Japan. Then there is the "cops-kicking-down-the-door" ending where interest rates rise but real rates are still negative. That's where the gold in the PP comes to the rescue.

All this raises a more basic question in my mind that we should answer before thinking about which asset allocation is best. How much of our surplus should we try to save as liquid assets? The world is awash in money, but the day of reckoning has not yet arrived. Maybe it is better not to save as much, but instead to try to pull forward expenses as much as feasible while money is cheap. For me that means focusing on paying off my mortgage quickly rather than trying to accumulate money for savings. Paying off debts pulls forward future consumption, lowers the functional APR on my note, and if I manage to pay off my debts before I become unemployed reduces my need for cash flow.

[sophie]

Another to consider is that a tight-money recession is only one of the possible ways this may end. I think of it as the "responsible-adults-in-charge" ending, but there is also the "party-on-Wayne" ending where bond prices remain high for the rest of your time on earth.

Maybe it is better not to save as much, but instead to try to pull forward expenses as much as feasible while money is cheap. For me that means focusing on paying off my mortgage quickly rather than trying to accumulate money for savings.

Meaning that we end up going the Japan route while the country/world slowly de-leverages.  Yes that's possible.  I hope not.  I'd rather suffer through the tight money recession.

I think a lot of people on this forum share your distaste for carrying debt.  There's an argument to be made for refinancing into a rock-bottom interest rate and then saving with the idea that your investment returns will be at least as good as the mortgage rate.  Then the money is there if you need it, and you can pay off the mortgage anytime you want.  With the fed being very anti-deflationary, I don't think that's a huge concern.

[One day at a time]

"I think a lot of people on this forum share your distaste for carrying debt.  There's an argument to be made for refinancing into a rock-bottom interest rate and then saving with the idea that your investment returns will be at least as good as the mortgage rate.  Then the money is there if you need it, and you can pay off the mortgage anytime you want.  With the fed being very anti-deflationary, I don't think that's a huge concern."

+1.  It may be that this is an economic moment when taking on suitable debt, ie. a mortgage, actually reduces risk going forward. 

[MachineGhost]

So I'm not dissuaded from the PP even in this nightmare scenario, although if long term bond rates get down to 1-1.5% I suspect there would be a lot of discussion about switching the bond allocation to cash or least intermediate term bonds.

I've already done that (drastically reduced the duration), but I'm not happy about it.  But I would be even more unhappy about it if I didn't do it and suffered a 20-year bear market in bonds.

The caveat is if you're going to kill the barbell, you have to use some kind of tactical allocation for the stocks because -- to paraphase Buffett -- they would be caught swimming naked when the tide goes back out.

[D1984]

Cash = No loss and a roughly 6.75% gain if you hold a savings account or a t-bill money market fund; roughly an 11.1% capital loss if you hald SHY (assuming a 1.75 year maturity and yields on it going from its current roughly 0.2% to 7.00%); the 11.1% capital loss would be offset by the increased yield so you would have a roughly 4.1% loss

In 1981, a 5-year ladder or single 3-year of T-Notes returned 15.73%.  If weighted appropriately, it was enough to reduce the losses in the other three assets by half.  The FFR at the time was 20% while the CPI was at 13.5%, so there is a real return gain even if it looks like a loss in nominal terms.

The 3-year T-Note was already yielding around 12.9% in late December 1980 and early January 1981 so it didn't experience much loss as rates rose (the 3-year rate never got quite as high as the FFR...it topped out at around 16.5% in the fall of '81 and went down from there); what do you think would happen to a 3-year note at today's yields (around 0.4%) if rates rose to 15% or even to, say, 7%?

Also, a 1 to 5 year ladder has two issues: One is that while you don't have any REALIZED losses unless you sell (hopefully you don't need to rebalance any amount more than what the amount that comes from your maturing one year rung of the ladder equals) you will still have a mark to market loss. Holding any treasury security to maturity pretty much guarantees you won't have a loss (at least not a nominal one) but in doing so you lose the ability to sell (for at least what you paid for the thing) because you have to hold it until maturity. The other issue is that while a ladder may prevent capital losses as above, it also means that you won't be getting into the higher yielding current issues at any real speed. Let's say you have a 1 to 5 year ladder at current rates, as follows:

1 Year Note = 0.14%
2 Year Note = 0.26%
3 Year Note = 0.42%
4 Year Note = 0.63%
5 Year Note = 0.79%

For an average yield of 0.45%. At the end of every year, the 1-year note matures and is redeemed, the other notes move down the maturity rung one year (since each is now a year closer to redemption), and you buy a 5-year note at the new prevailing rate with the proceeds from the 1-year note.

Now, let's say yields spike to 10% in a tightening shock. While you will replace the current five-year (said currently held five-year will become a four-year, the four-year will become a three-year, etc) with a new one with proceeds from the redeemed one-year as above, even if the new five-year yields 10%, the total average yield of your "cash" will still only be 2.42% ( because (0.26% + 0.42% + 0.63% +0.79% + 10.00% ) / 5 = 2.42% ); 2.42% isn't really enough to offset any serious loss in the rest of your portfolio. Now, you COULD sell securities of any maturity from two to five years to buy the new, higher yielding ones but what you will gain in yield on the new ones you will lose in capital value on the old ones. TANSTAAFL.

Holding 1 to 3 year treasuries in lieu of cash (or in lieu of t-bills) works out really well in a a generally falling rate environment (like the one we've had the past 31 years). It doesn't look so hot when rates are on a long-term uptrend.

[MachineGhost]

Holding 1 to 3 year treasuries in lieu of cash (or in lieu of t-bills) works out really well in a a generally falling rate environment (like the one we've had the past 31 years). It doesn't look so hot when rates are on a long-term uptrend.

What do you suggest then?

[D1984]

Holding 1 to 3 year treasuries in lieu of cash (or in lieu of t-bills) works out really well in a a generally falling rate environment (like the one we've had the past 31 years). It doesn't look so hot when rates are on a long-term uptrend.

What do you suggest then?

I dunno...I'm just here to poke holes in other people's theories, not to suggest any of my own. 

Seriously, I would probably say that at some point (based on interest rates...just like if T-bonds went to, say, 1.5% I think most of us would choose to hold more of the Treasury barbell as cash and less as LTTs than the typical 50/50 split) one has to make a choice to hold cash instead of 1 to 3 year Treasuries. I would probably say go with CDs with little or no withdrawal penalty but stay well under the $250K limit per bank (heck, stay well under the $100K limit if you're paranoid...and be sure to regularly check on the bank's financial strength and its capital ratios) or with a stable value fund if you are doing a PP in a 401K and it has such a fund.

Once (which may be a while) in the future rates rise back up somewhat, I might say switch some of the cash (or t-bills, or CDs, or other zero duration instruments) back to STTs but right now the risk outweighs the reward IMO.

[MachineGhost]

D1984,

We have 1/1942 to 6/1981 to look at to see the effect of rising interest rates and inflation on Treasuries:

T-Bills: 4.29% CAGR
5yr Ladder: 5.38% CAGR

Or 1/1962 to 6/1981 (10yr):

T-Bills: 6.21% CAGR
5yr Ladder: 6.93% CAGR
5yr Treasury: 6.95% CAGR
10yr Treasury: 7.01% CAGR

Or 1/1965 to 6/1981 (Inflation):

T-Bills: 6.76% CAGR
5yr Ladder: 7.49% CAGR
5yr Treasury: 7.51% CAGR
10yr Treasury: 7.56% CAGR

Or 8/1971 to 6/1981 (Nixon Shock):

T-Bills: 7.72% CAGR
5yr Ladder: 8.56% CAGR
5yr Treasury: 8.53% CAGR
10yr Treasury: 8.68% CAGR

Or 05/1972 to 6/1981 (30yr):

T-Bills: 8.00% CAGR
5yr Ladder: 8.76% CAGR
5yr Treasury: 8.72% CAGR
10yr Treasury: 8.86% CAGR
30yr Treasury: 3.31% CAGR

Seems to me, the longer durations do better than cash under both rising inflation and a fiat monetary system.  Its important to note that in reality most of us would be holding 30-year I-Bonds and 5-year CD ladders that won't flunctuate compared to Treasuries.  Still, I'm unclear where the upper limit is as theres not enough data to test 20 year maturities.  But even a 10-year maturity looks safe.  Unless I'm overlooking something, this really changes things.

EDIT: In the PP context, a 5yr ladder and a 5yr Treasury are virtually indistinguishable from each other except for a wee bit higher reward and risk for the latter.  Longer maturities definitely do not benefit the PP.  So given the need for liquidity when rebalancing, a 5yr ladder still seems best to me.

[D1984]

D1984,

We have 1/1942 to 6/1981 to look at to see the effect of rising interest rates and inflation on Treasuries:

T-Bills: 4.29% CAGR
5yr Ladder: 5.38% CAGR

Or 1/1962 to 6/1981 (10yr):

T-Bills: 6.21% CAGR
5yr Ladder: 6.93% CAGR
5yr Treasury: 6.95% CAGR
10yr Treasury: 7.01% CAGR

Or 1/1965 to 6/1981 (Inflation):

T-Bills: 6.76% CAGR
5yr Ladder: 7.49% CAGR
5yr Treasury: 7.51% CAGR
10yr Treasury: 7.56% CAGR

Or 8/1971 to 6/1981 (Nixon Shock):

T-Bills: 7.72% CAGR
5yr Ladder: 8.56% CAGR
5yr Treasury: 8.53% CAGR
10yr Treasury: 8.68% CAGR

Or 05/1972 to 6/1981 (30yr):

T-Bills: 8.00% CAGR
5yr Ladder: 8.76% CAGR
5yr Treasury: 8.72% CAGR
10yr Treasury: 8.86% CAGR
30yr Treasury: 3.31% CAGR

Seems to me, the longer durations do better than cash under both rising inflation and a fiat monetary system.  Its important to note that in reality most of us would be holding 30-year I-Bonds and 5-year CD ladders that won't flunctuate compared to Treasuries.  Still, I'm unclear where the upper limit is as theres not enough data to test 20 year maturities.  But even a 10-year maturity looks safe.  Unless I'm overlooking something, this really changes things.

EDIT: In the PP context, a 5yr ladder and a 5yr Treasury are virtually indistinguishable from each other except for a wee bit higher reward and risk for the latter.  Longer maturities definitely do not benefit the PP.  So given the need for liquidity when rebalancing, a 5yr ladder still seems best to me.

How did t-bills/cash do vs a 5-year ladder (or vs holding a 3 year Treasury) from 1945 to the early 1960s? That is how far back we have to go to when rates were as low as they are now. I noticed that none of your "fine grained" comparisons that included only 10 or 15 years (as compared to the "coarse grained" one from 1942 to 1981 which includes 40 years of widely varying inflation and interest rate conditions) started any earlier than 1962? I suspect a ladder (or a 3 year, or a 5 year) may have done worse from Jan 1st 1946 to to, say, January 1 1960.

Even if it DID do better, though, the question remains...how will a ladder do in a true "quick rising rate" shock vs the t-bills especially when starting from today's rather low rates? Rates took nearly 40 years to rise from their lows in the mid 1940s to their 1981 high....what will happen to a 5-year ladder (or a 1-3 year T-note fund like SHY) if that rise happens in four or seven or even ten years?

[craigr]

So I'm not dissuaded from the PP even in this nightmare scenario, although if long term bond rates get down to 1-1.5% I suspect there would be a lot of discussion about switching the bond allocation to cash or least intermediate term bonds.

My rough internal tripwire for LT bonds is around 2% or less. At that point I'm going to want to re-analyze the situation.

As for the cash buffer in a tight month recession. There will likely be nowhere to hide. I generally recommend that people hold around a year or two of living expenses and then optionally they can extend maturities to the 1-3 year range if they feel like. This is because you want the ability to wait out any drops in NAV of a fund so your cash in near term can buffer you until you need to tap the 1-3 year reserves. But this situation is likely only applicable to someone retired and, this is totally optional.

I have noticed at times that the 1-3 year range is very much a sweet spot in terms of how the market prices the asset as a cash substitute vs. going shorter. They are demanding more in interest due to the uncertainty, but it's not so far out on the time horizon that it gets really foggy like 5+ year notes/bonds.

Of course during a huge deflationary shock all bets are off. You better have LT bonds in the portfolio.

[MachineGhost]

How did t-bills/cash do vs a 5-year ladder (or vs holding a 3 year Treasury) from 1945 to the early 1960s? That is how far back we have to go to when rates were as low as they are now. I noticed that none of your "fine grained" comparisons that included only 10 or 15 years (as compared to the "coarse grained" one from 1942 to 1981 which includes 40 years of widely varying inflation and interest rate conditions) started any earlier than 1962? I suspect a ladder (or a 3 year, or a 5 year) may have done worse from Jan 1st 1946 to to, say, January 1 1960.

01/1945 - 12/1961:

T-Bills: 1.82% CAGR
5yr Ladder: 2.56% CAGR

Even if it DID do better, though, the question remains...how will a ladder do in a true "quick rising rate" shock vs the t-bills especially when starting from today's rather low rates? Rates took nearly 40 years to rise from their lows in the mid 1940s to their 1981 high....what will happen to a 5-year ladder (or a 1-3 year T-note fund like SHY) if that rise happens in four or seven or even ten years?

You aren't convinced that 1978 to 1981 was a "quick rising rate" shock?

[D1984]

You aren't convinced that 1978 to 1981 was a "quick rising rate" shock?

I believe I specifically said "a quick rising rate shock from our current low rates". 3-year T-note rates in early 1978 were already at around 7.40%; 3-year rates hit their highest at around 16.50% in late 1981. What do you think a rise of that magnitude (9.1 percentage points) would do to a three-year Treasury paying today's pathetic coupon rate of less than 0.5%? Even assuming these rise was over four years and not just over one year, it would certainly sting worse than the same 4 year rise when rates are already at the levels they were in 1978...I'm curious; what would happen to a 3 -year T-note (bought at three years and sold at two every year) if rates rose 2.275 percentage points each year and the yield on the two year and the three year maintained about the same gap (about 0.2 percentage points) they do now?

Bottom line, as long as even five year T-notes offer less than 1%, I'm sticking (in taxable accounts...for nontaxable it's the 401Ks stable value fund) mostly with savings accounts and rewards checking (all under FDIC and in fact under $25K per account) that offer anywhere from 0.80% to 3.00% with a duration of basically zero. When rates rise to maybe 2% across the 1-5 year ladder then perhaps I'll get back fully into STTs again.

P.S. where are you getting four year rates from (for the first year of your ladder...after that the five year becomes a four year and so on)...I'm having to extrapolate based on 3-year and 5-year yields?