Swedroe: Maturity of Fixed-Income Assets and Portfolio Risk

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply

User avatar
Sam2
Posts: 295
Joined: Thu Jun 14, 2007 6:31 pm

Post by Sam2 » Fri Dec 18, 2009 1:57 pm

It requires subscription.

Sam

User avatar
BlueEars
Posts: 3485
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Post by BlueEars » Fri Dec 18, 2009 6:29 pm

Cost is $45 to read this paper :( .

User avatar
Robert T
Posts: 2432
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Post by Robert T » Fri Dec 18, 2009 9:50 pm

.
Larry mentioned this work in an earlier thread(s). Here's a cut and past from one:

larry swedroe wrote:One of my coworkers and I just finished a research project on this very issue. We are submitting the paper to one of the Journals (I had Martin Fridson review and he thought we should submit it).

The findings were very interesting. Basically, the answer is it depends on the equity allocation. The higher the equity allocation the longer the maturity you can go--out to about intermediate is best, like 5 years or so. The lower the equity allocation the shorter the maturity should be, like very short.

FWIW – I got similar results - so agree with his findings.

Robert
.

User avatar
grabiner
Advisory Board
Posts: 20550
Joined: Tue Feb 20, 2007 11:58 pm
Location: Columbia, MD

Post by grabiner » Sat Dec 19, 2009 12:47 am

Robert T wrote:.
Larry mentioned this work in an earlier thread(s). Here's a cut and past from one:

larry swedroe wrote:One of my coworkers and I just finished a research project on this very issue. We are submitting the paper to one of the Journals (I had Martin Fridson review and he thought we should submit it).

The findings were very interesting. Basically, the answer is it depends on the equity allocation. The higher the equity allocation the longer the maturity you can go--out to about intermediate is best, like 5 years or so. The lower the equity allocation the shorter the maturity should be, like very short.

FWIW – I got similar results - so agree with his findings.


This is interesting. Intuitively, I would expect the opposite effect; the returns of long-term bonds are correlated with the returns of stocks, and therefore long-term bonds give less diversification benefit than short-term bonds to a stock-heavy portfolio. However, I'll trust these results, at least given then historical data. (The existence of TIPS may change things significantly.)
David Grabiner

User avatar
Adrian Nenu
Posts: 5228
Joined: Thu Apr 12, 2007 6:27 pm

Post by Adrian Nenu » Sat Dec 19, 2009 1:04 am

Depends what you use the bond allocation for:

Short duration bonds are better equity and inflation diversifiers than intermediate. Shorter duration bonds might be better suited for investors in the accumulation stage who have higher equity allocations than retirees. The period during the 70s & early 80s are a good example: high inflation and low equity returns. However, short duration bonds kept up with inflation while intermediate and long bonds got killed. TIPS would have obviously been even better equity diversifiers had they existed back then. So a combination of short duration high credit quality bonds, TIPS and FDIC insured CDs is the answer for the bond allocations of those investors who need equity diversification.

Intermediate duration bonds generally provide higher yield and more income but are not as good equity diversifiers as short duration bonds. Add TIPS for inflation protection. This might be more suitable for retirees who need reliable income streams and have relatively low equity allocations.

Adrian
anenu@tampabay.rr.com

conundrum
Posts: 813
Joined: Sat May 09, 2009 7:00 pm

Post by conundrum » Sat Dec 19, 2009 10:41 am

Robert T

Thanks for the link back to that November 2008 thread. For those interested it is a very informative thread with lots of good stuff from Larry, Robert T, Small Hi etc. Especially interesting because Larry has comments about his paper noted on this thread. Worth rereading the whole thread.

Drum

User avatar
Robert T
Posts: 2432
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Post by Robert T » Sat Dec 19, 2009 10:52 pm

grabiner wrote:This is interesting. Intuitively, I would expect the opposite effect; the returns of long-term bonds are correlated with the returns of stocks, and therefore long-term bonds give less diversification benefit than short-term bonds to a stock-heavy portfolio.

To try to unpack it: (references to bonds=US treasuries)

In ‘normal circumstance’ (periods which seem to drive the overall correlations): Average correlations between longer-term bonds and equities is higher than between short-term bonds and equities. If interest rates fall bond prices rise and stock prices also tend to rise as investors use lower discount rates on future company earnings (the reverse when rates rise). Hence the positive relationship, being higher for long-term bond with more interest rate sensitivity.

In periods of unexpected deflation (e.g. Great Depression) bonds perform well (as the real value of coupon payments rise), stocks perform poorly (earnings pressures with declining demand). Longer-term bonds tend to do better than short-term bonds, reflecting the higher real value of the longer stream of coupon payments). In this case correlations between stocks and bonds is negative, and likely more negative for stocks and longer-term bonds.

Code: Select all

Sept 1929-June 1932:  Return (%)
--------------------------------
Long-term treasury        14.3
Intermediate treasury     12.5
T-bills                    6.0
Total market             -83.3

In periods of financial crisis (e.g. 1998 Asian Crisis, 2008 credit crisis). Bonds perform well (flight to quality), stocks perform poorly. Historical evidence indicates longer term bonds do better than short-term bonds. Perhaps in anticipation of Fed cuts to stimulate growth (which would raise prices more on longer-term bonds than short-term bonds). Again in this case correlations between stocks and bonds is negative, and likely more negative for stocks and longer-term bonds.

Code: Select all

April to Sept 1998:           Return (%)
----------------------------------------
Vanguard Long-term treasury        12.7
Vanguard Intermediate treasury      9.8
Vanguard ST-treasury                5.5
Vanguard TSM                      -10.5   
Vanguard EM                       -35.4

2008                          Return (%)
----------------------------------------
LB 20+yrs Treasury                 33.7
LB 3-7yr Treasury                  13.3
LB 1-3yr Treasury                   6.7
MSCI US                           -37.0   
MSCI EAFE                         -43.4
MSCI EM                           -53.3               

In periods of unexpected inflation (e.g. 1970s). Stocks may in the short-term perform poorly, but should react positively in the longer-term (particularly small cap and value stocks). Bonds perform poorly as the real value of coupon payments declines. Longer-term bond would be more negatively impacted than short-term bonds. In this case the correlation between stocks and bond may initially be positive (when stocks do poorly) and more positive between stocks and long-term bonds. Some facts from the period of high US inflation 1973-1981 (in 1972 inflation was 3.4% and in 1982 was back down to 3.8%). The inflation protection over this period was provided by value and small cap stocks, not short-term bonds, and intermediate bonds were not far off short-term bonds.

Code: Select all

    Inflation   Large  Midcap    Small   LT Gov  Intermediate
                Caps   Value      Cap    Bonds      Bonds     t-bills

1973    8.8    -14.7    -15.6    -30.9    -1.1      4.6         6.9
1974   12.2    -26.5    -21.1    -20.0     4.4      5.7         8.0
1975    7.0     37.2     60.5     52.8     9.2      7.8         5.8
1976    4.8     23.8     48.6     57.4    16.8     12.9         5.1
1977    6.8     -7.2      7.3     25.4    -0.7      1.4         5.1
1978    9.0      6.6      9.1     23.5    -1.2      3.5         7.2
1979   13.3     18.4     28.2     43.5    -1.2      4.1        10.4
1980   12.4     32.4     17.3     39.9    -4.0      3.9        11.2
1981    8.9     -4.9     11.3     13.9     1.9      9.5        14.7
                        
AR      9.2      5.2     13.5     18.8     2.5      5.9         8.2

AR=annualized return
Source: Ibbotson Yearbook 2006.

The message seems to be look beyond average correlations. Longer-term bonds seem to have provided the greatest downside protection during deflation and financial crises, and small cap and value stocks provided the greatest protection during periods of high inflation. Obviously no guarantees.

Robert
.

Roy
Posts: 970
Joined: Wed Sep 10, 2008 9:34 am

Post by Roy » Sun Dec 20, 2009 8:49 am

Robert T wrote:The message seems to be look beyond average correlations. Longer-term bonds seem to have provided the greatest downside protection during deflation and financial crises, and small cap and value stocks provided the greatest protection during periods of high inflation. Obviously no guarantees.
Robert
.


The finding seems to indicate this is the better strategy when stocks have heavier weight in the portfolio since the addition of a small amount of bonds, even ST, does not seem to have much greater dampening effect than longer term bonds (the equity beta dominates). With a low Beta strategy, shorter is better. So, a low Beta "Larry" strategy that focuses on Small Value, uses ST Treasuries (TIPS aside), and might be better during inflation but not so good during deflation. I don't think going longer than 5 years is advised, which is not what most consider LT Bonds.

Roy

linuxizer
Posts: 1553
Joined: Wed Jan 02, 2008 7:55 am

Post by linuxizer » Sun Dec 20, 2009 10:29 am

Robert T wrote:Longer-term bonds seem to have provided the greatest downside protection during deflation and financial crises, and small cap and value stocks provided the greatest protection during periods of high inflation. Obviously no guarantees.


This is essentially Swensen's argument for LT Treasuries--you want as few of them as possible to help you survive deflationary and financial crises, when they are the only thing that rises. I believe he uses the term "insurance," which makes a certain amount of sense...costs you a premium most of the time but in a crisis it pays off.

User avatar
BlueEars
Posts: 3485
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Post by BlueEars » Sun Dec 20, 2009 12:07 pm

In Bernstein's recent book, The Investor's Manifesto (forward written July 09), he presents some portfolios and the bond sections are:

Code: Select all

all assets in retirement accounts
25%  VG Short-term Investment Grade (IG)
20%  VG Inflation-Protected
5%   money market

all assets in taxable accounts
15% Prime Money Market (later individual Treasury Notes)
15% Calif. Intermediate-Term Tax-Exempt
15% Limited-term Tax-Exempt
15% Short-term IG

assets in both taxable and retirement:
20% VG Short-term IG
15% VG Limited-term tax-exempt
10% VG Ohio tax-exempt
5%  money market

So right now he does not like Treasurys. In general he recommends "to keep it short, and keep it high quality" as he sees inflation as "the greatest single threat to any bond portfolio".

Short term IG bonds and Short-term Treasury did well in 3 crisis I've looked at: outbreak of Iraq war (Aug 02 - Apr 03), 911 attack (Sept 10 - Oct 1 2001), LTCM crash (Aug - Dec 1998). Short term IG didn't do well during the worst of the deflation fears in late 2008 but has made up most of the lost ground in 2009.

P.S. Thanks to Roger for the recent data he presents. I was thinking of doing something like that myself.

User avatar
Robert T
Posts: 2432
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Post by Robert T » Sun Dec 20, 2009 9:26 pm

.
linuxier,

Yes, the earlier post was along this lines of what Swensen described in his 1998 Pioneering Portfolio Management book (correlation section – pg 118). In addition, Larry’s 2003 The Successful Investor Today book has a section on The Value Premium, Inflation, and Portfolio Risk (pg. 238-9). Here’s an extract.

    "If you are going to hold a high allocation to value stocks, you should also consider holding longer-term fixed-income instruments. The recession/ deflation/ flight to quality risks of value stocks have negative correlation with the risks of longer-term fixed income instruments (which generally perform well in these types of environments). And since the evidence suggests that value stocks provide greater returns than growth stocks during inflationary periods, they offer greater protection than growth stocks against inflation risk inherent in longer term bonds. Thus holding a value tilt not only provides greater expected returns on your equity holdings, but it also allows you to take more interest-rate risk and thus potentially earn greater returns on your fixed income investments."
“Keep it short” seems to be the common advice (certainly from DFA, who’s first bond fund, if I recall, was a 1yr bond fund) and vigorously argued by some on the forum. But as far as I could tell from my analysis, the historical data didn’t suggest this was the optimal duration (at least for my 75:25 stock:bond allocation), and together with Swensen’s and Larry’s books – I have had a 0.5 term exposure target (i.e. intermediate term not short-term) for the last 7 yrs (and no credit exposure in fixed income). Obviously no guarantees.

Robert
.

User avatar
Robert T
Posts: 2432
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Post by Robert T » Tue Dec 22, 2009 8:18 am

.
Decomposing the Yield Curve. Interesting paper - seems to give similar explanations for the observed higher returns (price increases) of longer-term bonds in times of financial crisis (and reasons for yield curve inversions). From the paper:

“Suppose, for example, that investors expect economic activity to slow at some point in the future. If investors expect that weakness to require policy easing in the medium term, they will mark down their projected path of future spot interest rates, lowering far-forward rates and causing the yield curve to flatten or even to invert....”

And from Greg Mankiw

Q: Why would the yield curve ever invert?

A: The yield curve inverts when bond investors expect short-term interest rates to fall. They are willing to hold long-term bonds, despite the lower current yield, because they are locking in the yield. In other words, current long rates reflect both current short rates and expected future short rates. When investors expect a significant decline in short rates, long rates will be below current short rates.

Seems to suggest longer-term bonds may provide a useful hedge for both financial crises and slower economic activity (rising equities risks with yield curve invesions - if we believe that to be the case) as generally observed in the historical data.

Robert
.

the fixer
Posts: 75
Joined: Sat Jul 25, 2009 6:19 pm

Stocks and Bonds

Post by the fixer » Wed Dec 23, 2009 12:23 pm

I haven't gotten around to reading the paper, but did look at responses as well as comments made in one of the linked conversations from about 12 months ago. I was interested to update the data submitted in that previous post to include the other side of the big LT bond run up, and here are the statistics on various portfolio allocations using different fixed income maturities pasted with updated risk/return info:

Return/Standard Deviation (7/63 - 11/09):

81/19 Equity/ST Bond = +12.0% / 14.4
80/20 Equity/Int'd Bond = +12.0% / 14.3
78/22 Equity/LT Bond = +12.0% / 14.4

62/38 Equity/ST Bond = +11.0% / 11.1
60/40 Equity/Int'd Bond = +11.0% / 11.0
58/42 Equity/LT Bond = +11.0% / 11.7

42/58 Equity/ST Bond = +9.8% / 7.7
40/60 Equity/Int'd Bond = +9.8% / 8.0
36/64 Equity/LT Bond = +9.8% / 9.8

A casual look at various stock/bond combos using short term bonds (1-5YR), intermediate term bonds (5YR), and long term bonds (20YR) provides very few serious conclusions. My take:

--there is no meaningful difference in portfolio risk or return to stretch yourself beyond short term bonds to intermediate maturities, but with long dated bonds, risk does begin to rise at lower equity levels
--regardless of your stock/bond allocation, the low risk nature of short term bonds allows you to keep the smallest % possible in fixed income to achieve a given level of volatility reduction

Do realize, this period (63-09) is biased against short term bonds as prevailing interest rates experienced a net decline from 63- today, benefiting longer dated maturities. Studied over a prolonged period of a net interest rate increase (not unlikely from these levels), and its probable that balanced portfolio with short term bonds would have a noticeable risk/return advantage.

Vanguard ST Bond Index is as good as it gets for most. Interesting to note, however, it appears a short term variable maturity approach with a global reach (ie. DFA 5YR Global, which buys bonds in the 1-5 year maturity range and attempts to "ride the yield curve") offers the best of both worlds: returns in line/higher than intermediate term bonds with volatility (and therefore inflation sensitivity) like a short term strategy:

12/90 - 11/09 (return/annualized SD)
DFA 5YR Global = +6.4% / 4.1
5YR T-Notes = +6.3% / 6.2
1-5YR T-Notes = +5.5% / 4.0

User avatar
Robert T
Posts: 2432
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Post by Robert T » Wed Dec 23, 2009 11:17 pm

.
Following the same format as the previous post. Here is the return/SD and selected downside periods from 1925 to 2005 for high equity %portfolios with different bond term exposure (0, 0.5, and 1 term loads). FWIW - I get a different result, but we have been around this before. Few observations (at least my take):

    - Extending duration to intermediate term reduced volatility and downside risk, but there was not much added benefit extending to long-term bonds (although adding 2008 would perhaps show more benefit).
    - Adding intermediate bonds allowed for as small an equity allocation as low risk (‘risk-free’) t-bills for a given level of volatility reduction, while giving up less portfolio return.
    - Both these observations would have been reinforced if the 2006-09 time period was added to the analysis.

Code: Select all

                            1927-2005
Asset allocation            Return/SD    1929-32   1937   1973-74  2000-02

87/13 CRSP1-10/T-bills       9.5%/17.6    -62.2   -30.0    -33.9    -32.4
80/20 CRSP1-10/T-bills       9.1%/16.4    -58.6   -27.8    -30.8    -29.7
80/20 CRSP1-10/5yr T-notes   9.5%/16.4    -57.5   -27.4    -31.4    -26.0
79/21 CRSP1-10/20yr bonds    9.5%/16.4    -56.9   -27.3    -32.0    -23.4


Robert

PS: Here's an update of the numbers to end 2008 - the results are more favorable to long-term bonds.

Code: Select all

                            1927-2008 
Asset allocation            Return/SD    1929-32   1937   1973-74  2000-02   2008

89/11 CRSP1-10/T-bills       8.9%/18.4    -63.5   -30.9    -35.1    -33.4   -32.9
80/20 CRSP1-10/5yr T-notes   8.9%/16.6    -57.5   -27.4    -31.4    -26.0   -27.1
77/23 CRSP1-10/T-bills       8.4%/15.9    -56.4   -26.5    -28.9    -28.1   -28.1
79/21 CRSP1-10/20yr bonds    8.9%/15.9    -54.8   -26.2    -30.7    -21.1   -20.7

.
Last edited by Robert T on Thu Dec 24, 2009 6:15 am, edited 1 time in total.

dumbmoney
Posts: 2243
Joined: Sun Mar 16, 2008 8:58 pm

Post by dumbmoney » Thu Dec 24, 2009 12:36 am

Returns from October 12, 2007 to March 9, 2009 (S&P 500 peak to trough):

ssga SPY (S&P 500 index): -55.1%

ishares SHV (1-12 month treasury): 2.2%
ishares SHY (1-3 year treasury): 7.5%
ishares IEI (3-7 year treasury): 14.4%
ishares IEF (7-10 year treasury): 18.0%
ishares TLH (10-20 year treasury): 17.5%
ishares TLT (20+ year treasury): 22.5%

ishares TIP (TIPS): 1.2%

(Got the numbers from Yahoo, hope they are correct).
I am pleased to report that the invisible forces of destruction have been unmasked, marking a turning point chapter when the fraudulent and speculative winds are cast into the inferno of extinction.

natureexplorer
Posts: 4191
Joined: Thu Sep 03, 2009 10:52 am
Location: Houston

Post by natureexplorer » Thu Dec 24, 2009 12:43 am

dumbmoney wrote:Returns from October 12, 2007 to March 9, 2009 (S&P 500 peak to trough):

ssga SPY (S&P 500 index): -55.1%

ishares SHV (1-12 month treasury): 2.2%
ishares SHY (1-3 year treasury): 7.5%
ishares IEI (3-7 year treasury): 14.4%
ishares IEF (7-10 year treasury): 18.0%
ishares TLH (10-20 year treasury): 17.5%
ishares TLT (20+ year treasury): 22.5%

ishares TIP (TIPS): 1.2%

(Got the numbers from Yahoo, hope they are correct).


Did you calculate these number or are they listed on Yahoo somewhere? Do they account for dividends and interest?

dumbmoney
Posts: 2243
Joined: Sun Mar 16, 2008 8:58 pm

Post by dumbmoney » Thu Dec 24, 2009 12:51 am

natureexplorer wrote:Did you calculate these number or are they listed on Yahoo somewhere? Do they account for dividends and interest?


I calculated by hand using "adjusted close" numbers (supposed to be adjusted for dividends and splits). Not sure they are correct.
I am pleased to report that the invisible forces of destruction have been unmasked, marking a turning point chapter when the fraudulent and speculative winds are cast into the inferno of extinction.

natureexplorer
Posts: 4191
Joined: Thu Sep 03, 2009 10:52 am
Location: Houston

Post by natureexplorer » Thu Dec 24, 2009 9:26 am

dumbmoney, thanks.

User avatar
BlueEars
Posts: 3485
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Post by BlueEars » Thu Dec 24, 2009 12:03 pm

Robert T wrote:...(snip)...

Code: Select all

                            1927-2008 
Asset allocation            Return/SD    1929-32   1937   1973-74  2000-02   2008

89/11 CRSP1-10/T-bills       8.9%/18.4    -63.5   -30.9    -35.1    -33.4   -32.9
80/20 CRSP1-10/5yr T-notes   8.9%/16.6    -57.5   -27.4    -31.4    -26.0   -27.1
77/23 CRSP1-10/T-bills       8.4%/15.9    -56.4   -26.5    -28.9    -28.1   -28.1
79/21 CRSP1-10/20yr bonds    8.9%/15.9    -54.8   -26.2    -30.7    -21.1   -20.7
.

Hi Robert, I have a few questions about this data. First, do you just use the yields on the 5yr Treasury to get the yearly return and assume basically a repurchase each year? Or does this take bond capital gains (due to rate changes) into account?

Second, I'm guessing you don't have long term 2yr Treasury data to include a line for that. Don't know how far back they go. Where did you get the Treasury data so far back? The Fed 5yr constant maturity data I've seen doesn't go back further then about 1953.

Finally, many of us using Vanguard offerings would select between VFISX (short term Treasury) and VFITX (intermediate). VFISX might do better with inflationary pressures over the short and intermediate terms (maybe 2-5 yrs). VFITX has done better in crisis like last year where the capital gains component came in handy. Both are somewhat actively managed I guess. VFITX has done much better over the last decade but that was a declining rate environment as "the fixer" pointed out above. My guess is VFISX (in addition to TIPS) would be better for retirees but not sure about those who have even longer horizons. What are your thoughts?

Also a few observations. Your crisis data I believe (correct me if I'm wrong) does not include inflation/deflation so others might want to consider how the data is changed with that addition, especially the 1929-32 and 1973-74 periods. Also the cap gains component of bonds can really help in the crisis periods like last year but I suppose could hurt if we had a sudden inflationary upward move.

the fixer
Posts: 75
Joined: Sat Jul 25, 2009 6:19 pm

More on ST bonds

Post by the fixer » Thu Dec 24, 2009 6:04 pm

Properly measured, there just doesn't seem to be any meaningful benefit to longer duration bonds. From the chart above, looking at typical 60/40 allocations and substituting ST bonds (1mo t-bills and 5YR bonds until 1963, 1YR bonds and 5YR bonds since) for cash (t-bills):

Code: Select all

                            1926-2009
Asset allocation            Return/SD    1929-32   1937   1973-74  2000-02   2008

61/39 S&P 500/1-5YR          8.3%/11.6    -40.3   -21.0    -19.3    -16.8   -19.7
60/40 S&P 500/5YR            8.5%/11.6    -38.4   -20.4    -19.6    -12.2   -17.0
60/40 S&P 500/20YR           8.7%/12.2    -35.7   -19.5    -20.5     -5.4    -9.4


As for the lower returns of the S&P 500/1-5YR mix relative to the 5YR mix, that can be entirely explained by the fact that, prior to 1963, we are forced to use a combo of 1 month t-bills and 5YR bonds instead of 1YR and 5YR bonds for our "1-5YR" series.

How much difference does that make? Since 1963, 1YR bonds have had a term load of 0.1, a default load of 0.07, and an alpha of 0.07. If we apply those loadings to the 26-62 term and default results, we find that 1YR bonds would have outpaced 1 month bills by about 1.1% per year. Making that adjustment to the short term bond return series would have brought the S&P 500/1-5YR allocation return up to about 8.5%, in line with the other allocations.

Clearly, since 2000, investors have received better downside protection during bear markets by using 5YR bonds or 20YR bonds, but a quick glance reveals this really wasn't the case from the 20s through the 90s. Most of the reason the S&P 500/1-5YR allocation lost a bit more in previous bear markets was because it had a slightly higher equity allocation. Remember, however, this also would have made the allocation more tax efficient -- so there is a trade off.

Finally, adding short term bonds to an equity allocation (along with tilting to small and value stocks) gives you better inflation sensitivity, and a more reliably stable element to the portfolio that can be used for short term withdrawals or emergency liquidation needs.

To be fair, I don't think its a very big deal, however. If you put most of your emphasis on downside protection and believe the 2000s is more reflective of the future than the 20s through 90s, then you can assume the risk of longer term bonds. If you want the best combination of asset class stability, risk reduction, and inflation sensitive returns (on the fixed side ex. ST TIPS), while holding the most tax efficient combo of stocks and bonds (the lowest % of bonds of the 3 allocations I looked at) then short term bonds are a better bet.

User avatar
BlueEars
Posts: 3485
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Re: More on ST bonds

Post by BlueEars » Thu Dec 24, 2009 6:58 pm

the fixer wrote:...(snip)...
Clearly, since 2000, investors have received better downside protection during bear markets by using 5YR bonds or 20YR bonds, but a quick glance reveals this really wasn't the case from the 20s through the 90s. Most of the reason the S&P 500/1-5YR allocation lost a bit more in previous bear markets was because it had a slightly higher equity allocation. Remember, however, this also would have made the allocation more tax efficient -- so there is a trade off.

Finally, adding short term bonds to an equity allocation (along with tilting to small and value stocks) gives you better inflation sensitivity, and a more reliably stable element to the portfolio that can be used for short term withdrawals or emergency liquidation needs.

To be fair, I don't think its a very big deal, however. If you put most of your emphasis on downside protection and believe the 2000s is more reflective of the future than the 20s through 90s, then you can assume the risk of longer term bonds. If you want the best combination of asset class stability, risk reduction, and inflation sensitive returns (on the fixed side ex. ST TIPS), while holding the most tax efficient combo of stocks and bonds (the lowest % of bonds of the 3 allocations I looked at) then short term bonds are a better bet.

Hi fixer, thanks for your comments. I may a bit of an outlier in that my equity allocation varies quite a bit and is all non-taxable. I think I need to consider the bonds more in isolation to understand things. When I look at VFISX vs VFITX I see that from 8/96 to 6/09 the CAGR=5.2% and 6.7% respectively. So it seems for this period you did get paid to extend out. I guess you are pointing out that the 1920-90's period would not show as favorable bond fund results. Do you have data for this longer period comparing short and intermediate funds in isolation (2yr and 5yr as per Vanguards funds)? That is, the same table above but just the bond portion?

User avatar
grabiner
Advisory Board
Posts: 20550
Joined: Tue Feb 20, 2007 11:58 pm
Location: Columbia, MD

If you have a large taxable account

Post by grabiner » Thu Dec 24, 2009 7:57 pm

the fixer wrote:regardless of your stock/bond allocation, the low risk nature of short term bonds allows you to keep the smallest % possible in fixed income to achieve a given level of volatility reduction


This gives an advantage to using short-term bonds if your account is mostly taxable; this allows you to hold more stock in your taxable account, and if you do have to hold bonds in taxable, short-term bonds lose less to taxes than long-term bonds.
David Grabiner

the fixer
Posts: 75
Joined: Sat Jul 25, 2009 6:19 pm

Post by the fixer » Thu Dec 24, 2009 8:05 pm

Les,

When I look at VFISX vs VFITX I see that from 8/96 to 6/09 the CAGR=5.2% and 6.7% respectively. So it seems for this period you did get paid to extend out. I guess you are pointing out that the 1920-90's period would not show as favorable bond fund results. Do you have data for this longer period comparing short and intermediate funds in isolation (2yr and 5yr as per Vanguards funds)? That is, the same table above but just the bond portion?


The best we can do is to go back to 1963 (the inception of the ML 1YR T-Note Index). From 1963-1999, a period with wild swings in interest rates but similar starting/ending values, below are the annualized returns:

Cash = +6.3%
2YR = +7.3%
5YR = 7.5%
20YR = +7.1%

(cash = 1mo t-bills, 2YR = 80% 1YR, 20% 5YR, 5YR and 20YR = associated Ibbotson indexes)

Also note the dreaded 66-81 high inflation returns:

Cash = +6.8%
2YR = +7.1%
5YR = 5.8%
20YR = +2.5%

User avatar
BlueEars
Posts: 3485
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Post by BlueEars » Thu Dec 24, 2009 10:36 pm

Fixer, thanks for the data. Seems to me based just on the 1963-99 period one would probably want to stick with the 2yr Treasury.

Just as an aside, Bernstein as I mentioned above recommends short term IG bonds right now. I did one study that moved between investment grade VFSTX and short term Treasury VFISX based on the most recent 8mo returns and evaluated the first day of the each month. Also one must stay in a traded fund for 3 months.

Code: Select all

For short term bonds, Jan 91 to July 09:
VFSTX  5.7%/0.7   (second figure is std dev)
VFISX  5.3%/0.5
switch 6.1%/0.6  (65% of time in investment grade, 0.8 trade/yr)

And for intermediate bonds from Apr 97 to July 09:
VFICX  6.1%/1.5
VFITX  6.7%/1.5
switch  7.3%/1.4  (51% of time in investment grade, 1.3 trades/yr)


P.S. Merry Xmas to all the Bogleheads reading this :D

User avatar
Robert T
Posts: 2432
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Post by Robert T » Thu Dec 24, 2009 10:43 pm

.
Les,

Les wrote:First, do you just use the yields on the 5yr Treasury to get the yearly return and assume basically a repurchase each year? Or does this take bond capital gains (due to rate changes) into account?

The 5yr T-note returns were from the Ibbotson Yearbook – it includes income return (yield), capital appreciation, and a reinvestment return (monthly income reinvested in the total return index in subsequent months in the year). The return reflects holding bonds for the calendar yr (i.e. basically repurchase every yr) to reflect consistent 5yr. For overlapping years the returns are not dramatically different from the Vanguard Intermediate Treasury fund.

Second, I'm guessing you don't have long term 2yr Treasury data to include a line for that. Don't know how far back they go. Where did you get the Treasury data so far back? The Fed 5yr constant maturity data I've seen doesn't go back further then about 1953.

I don’t have the data.

Finally, many of us using Vanguard offerings would select between VFISX (short term Treasury) and VFITX (intermediate). VFISX might do better with inflationary pressures over the short and intermediate terms (maybe 2-5 yrs). VFITX has done better in crisis like last year where the capital gains component came in handy. Both are somewhat actively managed I guess. VFITX has done much better over the last decade but that was a declining rate environment as "the fixer" pointed out above. My guess is VFISX (in addition to TIPS) would be better for retirees but not sure about those who have even longer horizons. What are your thoughts?

Personally, I prefer VFITX for accumulation stage (for 75:25 stock:bond – where most of my analysis has focused), but prefer iShares BC 3-7yr Treasury to the Vanguard fund. For retiree’s not totally sure yet – but if you believe Larry’s Journal article, then for low equity allocations – shorter may be better (which is what I also found in a quick earlier look at it – linked to an earlier post).

Also a few observations. Your crisis data I believe (correct me if I'm wrong) does not include inflation/deflation so others might want to consider how the data is changed with that addition, especially the 1929-32 and 1973-74 periods. Also the cap gains component of bonds can really help in the crisis periods like last year but I suppose could hurt if we had a sudden inflationary upward move.

The earlier data includes the 1929-32 and 1973-74 periods. Yes unexpected inflation would hurt longer-term bonds relative to short-term bond. But even t-bills had negative real returns during the high inflation period (1973-1981). The only positive real returns US asset classes over that period were small cap and value stocks (again according to the data in the Ibbotson Yearbook).

Robert
.

User avatar
Robert T
Posts: 2432
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Post by Robert T » Thu Dec 24, 2009 10:59 pm

.
FWIW - I just get a different result to the fixer.

On the correct measure of short-term bonds:

    (i) The argument that is being put forward (my reading) is that the lower the risk in bonds the smaller the fixed income allocation needed to achieve a given level of volatility reduction. This seems to suggest that t-bills (the ‘risk-free’) fixed income should give even better results than short-term bonds (requiring even a lower fixed income allocation), and certainly outperforming 5yr T-notes and 20 yr government bonds in this regard. The data presented above indicates that historically this was not the case.

    (ii) On short-term (1-5yr) bonds – the best we can do is simulate a series back to 1927 as Ibbotson doesn’t publish such a series. So the question is how best to simulate a series. We have data on 1-5yr Government bonds from 1977. From 1977 to 2008 a 75:25 5yr T-note:T-bill mix had the same return but higher SD, and a 65:25 5yr T-note:T-bill mix had the same SD but slightly lower return.

Code: Select all

1977-2008
                               Return/SD

1-5yr*                           7.7/4.9
75/25 5yr T-notes/T-bills        7.7/5.4     
65/35 5yr T-notes/T-bills        7.5/4.9
      
*1977-1993 LB 1-5yr Gov (from Principia). 1994-2008 BC 1-5yr Trsy (from Vanguard).


I have added two short-term (1-5 yr) bond series to the table in my earlier post. The first short-term bond series (1-5yr (1)) is 75:25 5yr T-notes/T-bills from 1927-1976, Lehman Brothers 1-5yr Gov from 1977 to 1993 (from Principia) and Barclays Capital 1-5yr Treasury from 1995 to 2008 (from Vanguard – the ST Treasury fund benchmark). The second uses the 65:35 5yr T-note/T-bills from 1927 to 1976 (and the same as the first from 1977 to 2008).

The result is no different from my earlier post – over the period 1927 to 2008, for a relatively high equity allocation, extending duration reduced downside risk for a given return (the flip-side being allowing for a smaller equity allocation for a given level of volatility). Not inconsistent with Swensen’s take (from his Pioneering Portfolio Management book) “Portfolio managers wishing to reduce the opportunity cost of holding fixed income assets might rely on a small allocation to a long duration portfolio, in essence buying the diversifying power of bonds on the cheap”

Code: Select all

                            1927-2008 
Asset allocation            Return/SD    1929-32   1937   1973-74  2000-02   2008

89/11 CRSP1-10/T-bills       8.9%/18.4    -63.5   -30.9    -35.1    -33.4   -32.9
84/16 CRSP1-10/1-5yr(2)      8.9%/17.4    -60.3   -29.0    -33.0    -29.6   -29.8
83/17 CRSP1-10/1-5yr(1)      8.9%/17.3    -59.9   -28.7    -32.8    -29.2   -29.5
80/20 CRSP1-10/5yr T-notes   8.9%/16.6    -57.5   -27.4    -31.4    -26.0   -27.1
79/21 CRSP1-10/20yr bonds    8.9%/15.9    -54.8   -26.2    -30.7    -21.1   -20.7


Robert
.

User avatar
BlueEars
Posts: 3485
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Post by BlueEars » Sat Dec 26, 2009 12:58 pm

I'm still trying to get a feel for my next few years strategy. Currently rates are very low and the spread between 2yr and 5yr is high. Here is some data showing past performance of short term Treasurys (VFISX) and Intermediate Treasurys (VFITX) with particular emphasis on rising rate environments.

Image

Still clear as mud but I'm leaning towards short term bonds because they've done better in recent years where rates were very low and heading into a rising period.

the fixer
Posts: 75
Joined: Sat Jul 25, 2009 6:19 pm

Post by the fixer » Sun Dec 27, 2009 7:38 pm

Final thoughts on ST bonds:

FWIW - I just get a different result to the fixer.


We do seem to get different results. Not a big deal I guess...partly an issue of longest time periods available. My main argument is simply that '1mo t-bills' are not a worthy proxy for "short term bonds", and even a combo of t-bills/5YR or 1YR/5YR hasn't done a very good job of capturing short term bond returns since the mid 70s, bringing into question the longer term conclusions derived from short term bond benchmarks that have been cobbled together from various cash/5YR combinations.

Consider this: the Lehman (now Barclays) 1-5YR Treasury Bond Index has been simulated back to 1/76. Here are the term loads (ie. interest rate exposure) of various bond indexes and combinations over that period:

1/76-11/09 Term Coefficients

1mo t-blls = 0.00
1YR T-Notes = 0.10
1-5YR Index = 0.27
5YR Index = 0.46
40% t-bills/60% 5YR bonds = 0.27
55% 1YR notes/45% 5YR bonds = 0.27


...and associated annualized returns:

1mo t-blls = +5.7%
1YR T-Notes = +6.6%
1-5YR Index = +7.6%
5YR Index = +7.9%
40% t-bills/60% 5YR bonds = +7.0%
55% 1YR notes/45% 5YR bonds = +7.2%


So both "simulated combinations" of ST bonds have noticeably understated their actual returns over this stretch.

Here are the annualized returns/standard deviations of various balanced allocations over this period:

60/40 Stock/5YR Index = +12.4 / 11.0
62/15/23 Stock/t-bills/5YR = +12.2 / 11.1
62/21/17 Stock/1YR/5YR = +12.3% / 11.1
62/38 Stock/1-5YR Index = +12.4$ / 11.1

So whereas a cobbled together t-bill/5YR allocation to simulate "short term bonds" would indicate a 20bp lower return over this period, the actual index (Lehman 1-5YR T-Note) produced the same returns with only 0.1 more volatility, and 2% less overall bond allocation.

Finally, I don't see any long term benefit to using longer term bonds to hedge "value risk". Looking at all rolling 12 month periods since 1976 (Lehman 1-5YR T-Note inception) when value has underperformed growth, here are the associated bond returns:

1-5YR = +7.3%
5YR = +7.2%

(when I look at the same statistics, instead measuring 12mo rolling bond returns when "equity risk", or TSM minus t-bills is negative, I find a slight advantage for 5YR over 1-5YR (8.5% vs. 8.1%) ).

I don't think either one of these results overwhelmingly favors 5YR over ST bonds.

Robert and I just come to different conclusions on this. He prefers Int'd/LT bonds, and I like 'em short. All are free to draw their own conclusions.

User avatar
Robert T
Posts: 2432
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Post by Robert T » Sun Dec 27, 2009 10:24 pm

.
On proxy’s for short-term bonds - FWIW I used returns and SD to match t-bill/5-yr combinations, not term factor loads. How did they perform against the actual LB 1-5yr gov. bond returns from 1977-93, and BC 1-5 yr Treasury from 1994-2008? Actually they were fairly close, particularly the 75:25 5yr T-note:t-bill combination (as in table below). The results/conclusions for 1977-2008 seems no different from the result for 1927-2008 presented in the earlier post (except for the more significant performance of long-term bonds in the 1977-2008 period).

Code: Select all

                                                 1977-2008 
Asset allocation                                 Return/SD     2000-02   2008

89/11 CRSP1-10/1-5yr [LB/BC series]               10.2%/15.7    -32.3   -32.1
89/11 CRSP1-10/1-5yr [75:25 5yr T-note:t-bill]    10.2%/15.6    -31.8   -31.7
90/10 CRSP1-10/1-5yr [65:35 5yr T-note:t-bill]    10.2%/15.9    -33.0   -32.7

96/4  CRSP1-10/T-bills                            10.2%/16.9    -36.4   -35.6
80/20 CRSP1-10/5yr T-notes                        10.2%/14.2    -26.0   -27.1


So as best as I can tell, the historical data seems to favor intermediate over short-term (at least for my high equity allocation portfolio), and there seems to be fundamental reasons why this was the case (Swensen’s 1998 book). Just a different take from the fixer. Obviously no guarantees. We each have to decide.

Robert
.

kencc
Posts: 146
Joined: Mon Dec 01, 2008 12:23 am

Post by kencc » Mon Dec 28, 2009 10:48 pm

I have been trying to get a better understanding of the risk (in terms of actual potential losses) of holding a portfolio with a high bond allocation. Although potential losses from holding stocks are well documented there appears to be little analysis of such risks arising from holding bonds during, for example, the bond bear market from 1946 to 1981.

From one of Gummy's spreadsheets I find the following:-

1946-1981 Annualized Returns.
........................ Nominal ..... After inflation
T bills (cash) ..... 4.1% ......... -0.5%
5 yr Treas ......... 3.8% ......... -0.8%
Long Bonds ....... 2.0% ......... -2.5%

1966-1981 Annualized Returns
........................ Nominal ..... After inflation
T bills (cash) ..... 6.8% ......... -0.1%
5 yr Treas ......... 5.8% ......... -1.2%
Long Bonds ....... 2.5% ......... -4.2%

Therefore in real terms, over the 35 year bear market it appears 5yr Treasuries lost about 25% and Long Bonds lost about 60%. My 'age-in-bonds' portfolio would be stocks/bonds 35/65 but I wouldn't feel any more comfortable with that after looking at the 1946/81 scenario than I would with 65/35. However I may have misunderstood the 1946/81 figures?

User avatar
BlueEars
Posts: 3485
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Post by BlueEars » Mon Dec 28, 2009 11:10 pm

Ken, you might want to look at the FIRECalc application to answer some of your questions. It's free and has switches for the sort of questions I think you are asking: http://firecalc.com/firecalc2.php I'd suggest using the advanced version.

kencc
Posts: 146
Joined: Mon Dec 01, 2008 12:23 am

Post by kencc » Tue Dec 29, 2009 9:42 am

Les, I've used firecalc before but it only shows historical coincidences of stock/bond performances. But I'm also interested in considering scenarios outside of those particular coincidences.

Following on from recent stock performance there seems to have been a lot of discussion about setting the equity allocation % without reference to bonds but in terms of max-tolerable-loss or sleep-well-factor or money-one-can't-afford-to-lose e.g. historically stocks have lost 50% therefore if portfolio losses should not exceed 10% then stocks/bonds=20%/80%. However, if I'm not mistaken, there have been a number of periods in the past when short duration bonds have lost around 30% and long bonds 60% in real terms. If one had 80% bonds then portfolio losses would be 25% to 40% in circumstances where stocks break even. I’m only doubtful about whether these figures are correct because I've never seen this sort of analysis anywhere else.

User avatar
BlueEars
Posts: 3485
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Post by BlueEars » Tue Dec 29, 2009 10:57 am

Ken, perhaps I do not understand your requirements but here is what I was thinking. Run FIRECalc in advanced mode requesting the spreadsheet results. Turn off things like other cash flows (social security, etc). Then try various combos of equities, LT bonds, 5yr Treasury, commercial paper. Look at various subperiods in the spreadsheet and compare the cases. You are not just interested in the end result but all the years in between.

kencc
Posts: 146
Joined: Mon Dec 01, 2008 12:23 am

Post by kencc » Wed Dec 30, 2009 2:00 am

Les, using a phrase Robert T mentioned in an earlier post, I'm trying to 'look beyond average correlations' and also, for this purpose, to ignore how particular stock/bond allocations may have behaved historically and to ignore considering risk in terms of standard deviations i.e. if I hold a certain amount of bonds how much potential loss might I have to anticipate from that part of my portfolio in money terms if a bear market in bonds occurs and would I find that loss unacceptable.

My information is limited but I'm more confident in my figures now which show from 1928 to 2000 there were two periods where holding bonds resulted in losses in real terms:-
................ loss from 1940 to 1959 .......... loss from 1971 to 1981
T bills ................... -40% ............................. -12%
5yr Treas .............. -35% ............................. -25%
Long bonds ........... -38% ............................. -42%

Which, on this basis, if I had a high bond allocation portfolio it confirms to me that I would go with short duration bonds. However if I compare the results against a bench mark of, say, the S&P500 which over the last 10 year period has lost about 30% in real terms (dividends reinvested) then I wonder why a high bond portfolio is generally considered to be "safer" than a high equity portfolio. I understand the concept of risk in terms of standard deviations and assessing stock/bond allocations accordingly, but losses from both bonds and equities over a 10 year or so period can be within a similar range of 25-35% in real terms. Therefore, from that perspective, it appears to me that high bond allocations are no "safer" than high equity allocations.

User avatar
BlueEars
Posts: 3485
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Post by BlueEars » Wed Dec 30, 2009 11:31 am

Hi Ken, I think I see better where you are going with this. I've not yet considered getting into a mostly bond portfolio. There is an Aug 2003 article by Bill Gross that gives good charts of past bond returns: http://www.pimco.com/LeftNav/Featured+Market+Commentary/IO/2003/IO_08_2003.htm Here is one table from that article:

Image

Up to the early 1980's intermediate Treasurys didn't look good at all. Gross refers to the 1930-1980 period as: the pathetic "suckers" negative real return :). TIPS and short bonds would seem to be a good answer for today's retirees leaning towards increased bond allocations and worried about future inflation.

User avatar
Robert T
Posts: 2432
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Post by Robert T » Fri Aug 20, 2010 7:44 pm

.
FWIW - just been looking at this again. Here are some additional results. For portfolio with highest Sharpe ratios - long-term bonds were more prominent as fixed income in portfolios with higher equity allocations and for portfolios with a higher SV tilt.

Code: Select all


1927-2005: Portfolios with highest Sharpe Ratios

Allocation (%)

Stock/       Equity    Bonds
Bond         ------    ------------------------------------------   
              SV       t-bills    5 yr T-Notes   20 yr Government

90/10         90         0            0             10 
80/20         80         0            0             20 
70/30         70         0            1             29
60/40         60         0           25             15
50/50         50         0           50              0   
40/60         40         0           60              0
30/70         30         0           70              0
20/80         20         0           80              0
10/80         10        14           76              0         

           
Stock/       Equity    Bond
Bond         ------    -----------------------------------------
          30% S&P500
          30% LV
          20% Small
          20% SV       t-bills    5 yr T-Notes   20 yr Government

90/10        90         0            0             10 
80/20        80         0            9             11 
70/30        70         0           30              0
60/40        60         0           40              0 
50/50        50         0           50              0   
40/60        40         0           60              0
30/70        30         0           70              0
20/80        20         0           80              0
10/90        10        10           80              0       


.

User avatar
Noobvestor
Posts: 4439
Joined: Mon Aug 23, 2010 1:09 am
Contact:

Post by Noobvestor » Wed Nov 24, 2010 5:33 pm

Not to read too much into this graph, but it seems to suggest that the tipping point for S&D tilted portfolios between intermediate and long is around the 80% mark, and for 'normal' lumper porfolios is around the 60-65% mark, but that shorter durations never really give you the ideal sharpe ratio. Given these long-term results, I'm really puzzled by the current trend toward tilting short on duration - I guess it's just a market-timing move, or perhaps I'm oversimplifying? I realize rates can't go down much, but that doesn't mean they have to go up, either.

the fixer wrote:To be fair, I don't think its a very big deal, however. If you put most of your emphasis on downside protection and believe the 2000s is more reflective of the future than the 20s through 90s, then you can assume the risk of longer term bonds. If you want the best combination of asset class stability, risk reduction, and inflation sensitive returns (on the fixed side ex. ST TIPS), while holding the most tax efficient combo of stocks and bonds (the lowest % of bonds of the 3 allocations I looked at) then short term bonds are a better bet.


Interesting conclusion. It seems to omit deflation protection though. If S/V are good for inflation, then couldn't the argument be made that an SV-tilted portfolio therefore needs more deflation hedging, via longer-term bonds (downside benefits of the 2000s aside)?
"In the absence of clarity, diversification is the only logical strategy" -= Larry Swedroe

larryswedroe
Posts: 15597
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

noob

Post by larryswedroe » Wed Nov 24, 2010 5:54 pm

The idea behind highly tilted portfolios is to use the tilt to have a low equity allocation, which lowers the need for deflation protection. And besides if you own TIPS you have both inflation and deflation protection, especially if you buy longer term TIPS and get the term premium without the inflation risk

User avatar
Noobvestor
Posts: 4439
Joined: Mon Aug 23, 2010 1:09 am
Contact:

Re: noob

Post by Noobvestor » Wed Nov 24, 2010 6:27 pm

larryswedroe wrote:The idea behind highly tilted portfolios is to use the tilt to have a low equity allocation, which lowers the need for deflation protection. And besides if you own TIPS you have both inflation and deflation protection, especially if you buy longer term TIPS and get the term premium without the inflation risk


Hi Larry - thanks so much for the response. Good point about the overall lower need for deflation protection re:higher bond allocation (+ the dual benefits of TIPS). I think that was the 'missing link' in this thread, explaining why an SV tilt with lower equities provides *both* inflation and deflation protection.
"In the absence of clarity, diversification is the only logical strategy" -= Larry Swedroe

gwayne101
Posts: 1
Joined: Tue Jun 28, 2011 2:37 am

Post by gwayne101 » Fri Jul 15, 2011 2:35 am

Code:


1927-2005: Portfolios with highest Sharpe Ratios

Allocation (%)

Stock/ Equity Bonds
Bond ------ ------------------------------------------
SV t-bills 5 yr T-Notes 20 yr Government

90/10 90 0 0 10
80/20 80 0 0 20
70/30 70 0 1 29
60/40 60 0 25 15
50/50 50 0 50 0
40/60 40 0 60 0
30/70 30 0 70 0
20/80 20 0 80 0
10/80 10 14 76 0


Stock/ Equity Bond
Bond ------ -----------------------------------------
30% S&P500
30% LV
20% Small
20% SV t-bills 5 yr T-Notes 20 yr Government

90/10 90 0 0 10
80/20 80 0 9 11
70/30 70 0 30 0
60/40 60 0 40 0
50/50 50 0 50 0
40/60 40 0 60 0
30/70 30 0 70 0
20/80 20 0 80 0
10/90 10 10 80 0


.whoah! I'm really bad with math, especially computations such as this one. Are you the one who computed this? And how did you know how to compute an insurance.

fishndoc
Posts: 2323
Joined: Wed Apr 11, 2007 11:50 am

Post by fishndoc » Fri Jul 15, 2011 7:16 am

Resurrections of threads like this remind me how much I miss Robert T's posts on this forum. :(
" Successful investing involves doing just a few things right, and avoiding serious mistakes." - J. Bogle

Bongleur
Posts: 2037
Joined: Fri Dec 03, 2010 10:36 am

Post by Bongleur » Sat Jul 16, 2011 11:01 pm

Robert T wrote:To try to unpack it: (references to bonds=US treasuries)

In ‘normal circumstance’ (periods which seem to drive the overall correlations):

In periods of unexpected deflation (e.g. Great Depression)

In periods of unexpected inflation (e.g. 1970s).

The message seems to be look beyond average correlations. Longer-term bonds seem to have provided the greatest downside protection during deflation and financial crises, and small cap and value stocks provided the greatest protection during periods of high inflation. Obviously no guarantees.

Robert
.


Is there such a thing as a period of "expected deflation?"

And "expected significant increase in sovereign risk premium?"

So what happened to RobertT ?
Seeking Iso-Elasticity. | Tax Loss Harvesting is an Asset Class. | A well-planned presentation creates a sense of urgency. If the prospect fails to act now, he will risk a loss of some sort.

TallyMan
Posts: 135
Joined: Wed Dec 26, 2007 10:04 pm

Post by TallyMan » Sun Jul 17, 2011 12:46 pm

Robert T wrote:

Code: Select all

    Inflation   Large  Midcap    Small   LT Gov  Intermediate
                Caps   Value      Cap    Bonds      Bonds     t-bills

1973    8.8    -14.7    -15.6    -30.9    -1.1      4.6         6.9
1974   12.2    -26.5    -21.1    -20.0     4.4      5.7         8.0
1975    7.0     37.2     60.5     52.8     9.2      7.8         5.8
1976    4.8     23.8     48.6     57.4    16.8     12.9         5.1
1977    6.8     -7.2      7.3     25.4    -0.7      1.4         5.1
1978    9.0      6.6      9.1     23.5    -1.2      3.5         7.2
1979   13.3     18.4     28.2     43.5    -1.2      4.1        10.4
1980   12.4     32.4     17.3     39.9    -4.0      3.9        11.2
1981    8.9     -4.9     11.3     13.9     1.9      9.5        14.7
                        
AR      9.2      5.2     13.5     18.8     2.5      5.9         8.2


AR=annualized return
.


Any thoughts on whether Large-Caps, Mid-Caps, and Small Caps would would behave the same now if we have rising interest rates and rising CPI akin to 1973-1981? Perhaps with the ability of Large Caps to make profits overseas, over the entire period, they would outpace SCs now? MCs as a "sweet spot" to take advantage of their "room to grow," perhaps with an ability to profit from overseas trade?

Thx for your thoughts,
Steve

Post Reply