Gyroscopic Investing

Some Forumers here have mentioned using EDV.  I'm aware of 2 uses for EDV:

1 For Vanguard Brokerage Svcs (VBS) customers in a tax-advantaged account.  To implement the 30 yr T-Bond asset of the PP, EDV is seemingly the only Vg bond fund that is 100% USTs (VUSTX has some non-UST bonds from issuers like Fannie Mae).

2 For PRPFX mutual fund users, PRPFX does not have enough Deflation/30 yr T-Bond protection, so some PRPFX users combine 90% PRPFX / 10% EDV in order to be closer to the orthodox HBPP asset alloc.

Now the orthodox HBPP approach recommends holding individual UST security(ies) directly, over holding an indirect 30 yr UST-Bond fund like TLT/VUSTX/EDV.  Currently, Fidelity & VBS both seem to offer commission-free purchasing of individual USTs with $1000 minimums (apparently the standard minimum for USTs anyways - would presumably be the same min at any broker).  So some Forumers have commented on replacing their indirect holding (eg TLT) with direct holding (specific individual UST-Bond(s) ).

Out of curiosity, has anyone done the same with EDV, in the sense of intentionally replacing the indirect EDV with a direct Zero Coupon UST Security?  If so, could you explain why/how you use your direct Zero Coupon USTs, & any advice/notes about using them?

Off hand, it seems that Zero Coupon's are much less liquid than T-Bonds, & have a huge bid-ask spread, which would make direct Zero Coupon usage prohibitively expensive.

--
quick bid-ask spread comparison:
Checking the Vanguard Bond Desk right now.  Put in 2042 for both the "From Maturity" & "To Maturity" fields.

There's 1 T-Bond, maturing 2042-May-15, bid price = $106.390625. ask price = $106.5.
bid ask spread pct (BASP) = 0.10% (106.5 - 106.390625) / 106.390625

otoh there are 4 Zero Coupon UST Securities, 2 maturing 2042-May-15 & 2 maturing 2042-Feb-15.  Here's there bid price/ask price, BASP:
41.947/42.826, BASP = 2.10%
42.279/42.876, 1.41%
42.037/42.687, 1.55%
41.337/42.838, 3.63%
-

Imho the bond market is far more complicated than the stock market.  There's very little intelligent commentary in the financial press and little online data I've been able to find.  That's why your post probably won't generate as many responses.

You bring up a good point about the bid/ask price on the zeros.  I noticed that when I leveraged up my long bond position when I bought some 30 year zeros.  I also noticed When I bought the zeros there was a spread of about 25 basis points between a 30 year nominal and a 30 year zero, a 40 bp spread on 20 year nominal/zero treasuries, and 10 bp spread on 10 year nominal/zero treasuries.  I assume (perhaps wrongly) that these spreads will narrow should yields drop considerably.  I did find a nugget of information online that spreads on the 30 year treasuries was in the 10 bp range a few years ago.  Today I see the spread between the 30 year nominal and 30 year zero is 18 bp.  So for me so far, the drop in the spread has more than offset the bid/ask spread of 1.5 percent I got on my zero bond.  Also, edv was selling for a 0.8 percent premium At the time so that would have to figure In as well in comparison. 

I should note that I didn't exchange edv for direct ownership of the zeros - I sold some stocks in my Ira and bought zeros.

Even though the bid/ask spread is rather hefty, I'm not sure that necessarily precludes buying the zero.  Of course with tighter spreads between the 30 year nominal and zero today, it may be better to go with the nominal.  It may be worth considering buying shorter duration zeros (20-25 years) if the spread is still In the 40 bp range.  Just a thought.

Sorry for the choppy writing - I'm forced to posting through my iPad.

Clive wrote: For less liquid, more volatile assets you might reasonably expect to see wider bid ask spreads.

This chart shows sorted daily (absolute) price changes for TLT and ZROZ over the last couple of years

[align=center][/align]

As you might expect, ZROZ is clearly more volatile than TLT, so likely commands a higher bid/ask spread.

You might like to resort to shorter term trading methods in such cases. Use limit orders and for buying set the price to the bid price and for selling set the price to the ask price. Also look to buy on down days (when there are more selling than buying) and sell on up days (when there are more buying than selling). That way you're providing liquidity to the market and are more likely to eliminate any bid/ask spread overheads entirely overall (and maybe even make a bit extra). If you're in more of a hurry to buy (sell), then set the limit price to the mid price. If you're desperate to buy (sell), then you just have to take the hit of the additional cost of the full bid/ask spread.

Doesn't Vanguard also do something like 'free' trading? If so I would suspect that their bid/ask prices might be wider than others. If you look around elsewhere you might find others with tighter spreads.

Sounds reasonable.  When I get back to the US I may set up a spreadsheet and try to quantify your idea of buying on down days And selling on up days. 

I appreciate the insight - Thanks!

FarmerD, thanks for your thoughts.

when you refer to basis points, I assume you are referring to yield to maturity, correct?

Are you using your 30 yr Zeros for implementing a PP asset alloc?  If so, do you use them in the orthodox PP method, in terms of buying the longest maturity zero available close to 30 yrs old, & barring a selling Bond rebalance, selling it 9-10 yrs later when it has 20 yrs left to maturity?
-

Edit: I know 1 basis point is 0.0001 or 0.01%.  FarmerD, it seems you're referring to the amount of bid-ask spread, expressed in basis points.  Sorry for my earlier confusion, thinking you referred to the slightly higher yield (at least at the time I saw the quote at VBS Bond Desk) on the 30 yr Zero vs 30 yr t-Bond.

cabronjames wrote: FarmerD, thanks for your thoughts.

when you refer to basis points, I assume you are referring to yield to maturity, correct?

Are you using your 30 yr Zeros for implementing a PP asset alloc?  If so, do you use them in the orthodox PP method, in terms of buying the longest maturity zero available close to 30 yrs old, & barring a selling Bond rebalance, selling it 9-10 yrs later when it has 20 yrs left to maturity?
-

Edit: I know 1 basis point is 0.0001 or 0.01%.  FarmerD, it seems you're referring to the amount of bid-ask spread, expressed in basis points.  Sorry for my earlier confusion, thinking you referred to the slightly higher yield (at least at the time I saw the quote at VBS Bond Desk) on the 30 yr Zero vs 30 yr t-Bond.

Actually I was referring to yield to maturity in my earlier post.  Yep...that what happens when you're suffering jet lag.  My writing gets choppy AND misleading.

i still think if the spread on the yield to maturity between the 30 year nominal and the 30 year zero is sufficiently large, that may offset the bid/ask spread on the zero.  I wish I could find more data on the historical ytm spread between zeros and nominals.

https://personal.vanguard.com/us/funds/bonds/bonddesk
fyi, checkin the Vanguard Bond Desk now (2012-Jul-12, ~3p ET), looks like some more reasonable bid-ask spreads (let me abbreviate as BAS) are available for individual Zero Coupons.

btw it appears the general trend within Zero Coupons, is for Principal payments seem to have a lower/better bid-ask spread than the Interest payments.  This seems intuitive to me, if my understanding of a UST-Bond's cash flow is correct, since presumably the Principal payment is always $100 (in nominal 2042 $s), whereas the Interest payment would be something like $1.50 (in nominal 2042 $s), (3% annual interest paid twice a year, every 6 months).  Since the dollar value of a Principal payment market is >10X more than that of an Interest payment market, it intuitively seems that the Principal market would have a lower bid-ask spread.

The Principal payments in 2042 still have a >1% bid-ask spread.

However, there are 2041 Principal with bid-ask spread under 1%.  The Principal maturing 2041-08-15 has a bid/ask of $45.224/$45.439, BAS = 0.48%.  The other 2 Principals maturing 2041-02-15 or 2041-05-15 were similar, at most 0.71%.

What about when the Zero hits 20 yrs left until maturity & thus it's time to sell the Zero, assuming using the same HB idea of buying a UST-Bond that has 30-29 yrs till maturity, & selling it 10-9 yrs later once it hits 20 yrs till maturity?  I checked the 2032 Zeros.  No Principal Zeros came up in the search, but 4 Interest payment Zeros did, with BAS bt 0.38%-0.48%.

afaict it seems like the risk of direct buying of a Zero is 9-10 yrs later & it's time to sell (or if you need to do a sell rebalance, as needed per the rebalance signal).  You can specifically only BUY a Zero that has a lower <0.75% BAS, but when you need to SELL it, perhaps there is risk that there's a bogus/high ~2% BAS instead of a <0.75% BAS.

--

Afaict, my quick calculation on the costs of UST Zero, indirect with EDV or direct Zeros, would say that with a <0.75% BAS, direct Zero has equal or better investing costs than indirect/EDV.:

Assumptions
1 I "assume the worst" in that you pay the entire BAS on both the Buy & Sell transaction.  This is based on my real historical experience in Buying & Selling EDV online with a limit order as a VBS (Vanguard Brokerage Svcs) customer.  To actually get the transaction to execute, I find that putting a price in the middle of the bid & ask will not execute.  I end up putting a limit price about about 75% mostly towards the opposite price (eg if I'm Selling the Bid price, if I'm Buying the Ask price).  I've never bought direct Bonds at the VBS Bond Desk, but I would rather similarly "assume the worst".  I don't think it's appropriate or realistic to assume you only pay half the BAS, assuming you can get a price at the midpoint in bt the bid & ask price.

2 I assume, per the BAS I found at the VBS Bond Desk, that it's cheaper to buy Zeros that mature in ~29 yrs, as opposed to the latest available ~30 yr Zero.  I also assume that no Sell Rebalance is required.  Hence, I calculate costs over a 9 yr cycle: Buy a ~29 yr old Zero, wait 9 yrs, & sell that Zero when it has ~20 yr left until maturity

3 I assume the BAS for buying & selling a direct Zero is 0.75%

4 I assume the BAS for EDV is 0.22%, based of the current 30 day avg BAS Vanguard self-reports here
https://advisors.vanguard.com/VGApp/iip ... daskspread

5 The current EDV annual expense ratio is 0.13%
--

Calculation:

Direct Zero cost over the 9 yr cycle = 2 * BAS = 2 * 0.75% = 1.50%

Indirect EDV cost over the 9 yr cycle = (2 * BAS) + (( (1+exp ratio) ^ 9) - 1) = (2 * 0.0022) + ((1.0013 ^ 9) - 1) = 1.62%

--

So, in summary the direct Zero has a BAS <0.75%, it would seem that it's investment costs are competitive with indirect/EDV.  However, the direct Zero method MIGHT face a risk of a <0.75% BAS not being available 9 yrs later when the Zero needs to be sold when it hits 20 yrs remaining until maturity.
--

I'd be interested in your feedback on this post.  Thx again & thx in advance!