American Funds Bold Claim

I'll try to keep this post as short as possible but it seems doubtful. Earlier this year investors were regaled with ads from American Funds claiming: "A Group of Select Equity Funds Has, on Average, Consistently Beaten the Index"

Key words here are "group" and "consistently". I guess this implies that individually and for some periods they don't beat the index. After all kinds of vague charts and unspecified data, the last page of their pdf gets specific.

Seven American Funds are represented as a select group that can beat the index (VFINX).

	Symbol	Inception
Investment Company of America A	AIVSX	1/2/1934
American Mutual Fund A	AMRMX	2/21/1950
Washington Mutual Investors Fund A	AWSHX	7/31/1952
AMCAP Fund A	AMCPX	5/1/1967
The Growth Fund of America A	AGTHX	11/30/1973
Fundamental Investors A	ANCFX	8/1/1978
The New Economy Fund A	ANEFX	12/1/1983

Vanguard S&P 500 Investor Fund

VFINX

8/31/1976

Now lets acknowledge that these are great American mutual funds. The oldest fund here, the Investment Company of America goes back to 1934! Not sure what funds, if any, merged into AIVSX (or the others); but that would have a direct bearing on survivor bias and invest-ability.

I always wondered if any of the old funds have shareholder records going back to 30 years or more. I wonder what those old accounts actually look like today.

To test American's claim I took Yahoo's dataset, and compared the above funds. Caveat: I am comparing NAVs to NAVs, fees NOT included.  Is this fair? Well, both American and Vanguard have different fund classes with differing fees. American Fund A shares have some of the highest loads in the business, but then again, they are legacy funds. So to avoid the load penalty, I am using NAV for comparisons.

And these are the results:

VAMI RETURNS

VAMI Table:

AWSHX   $20,806.88

AGTHX   $22,019.03

ANCFX   $21,394.35

AIVSX    $14,844.63

AMRMX  $13,950.99

AMCPX   $12,828.53

ANEFX   $13,841.07

VFINX   $22,008.00

Only three funds, AWSHX, AGHX and ANCFX compare favorably to VFINX over the long term.  For shorter periods, nearly ALL the listed American Funds do well.

ROLLING RETURNS

Q32017
	VFINX	AWSHX	AGTHX	ANCFX	AIVSX	AMRMX	AMCPX	ANEFX
YTD	13.21%	12.14%	16.54%	14.56%	12.02%	11.34%	14.10%	24.17%
1-Year	16.93%	17.63%	18.11%	18.48%	15.05%	14.41%	15.83%	23.33%
3-Year	10.13%	9.39%	10.94%	10.88%	8.87%	9.09%	9.06%	10.79%
5-Year	13.15%	12.71%	14.10%	13.61%	12.89%	11.81%	13.64%	14.80%
7-Year	13.28%	12.85%	13.10%	12.76%	12.10%	11.71%	13.18%	13.50%
10-Year	7.06%	6.62%	7.09%	7.01%	6.35%	6.84%	7.70%	7.92%
15-Year	9.45%	9.06%	10.27%	10.67%	8.98%	8.97%	9.67%	11.43%
20-Year	6.67%	7.30%	9.05%	8.22%	7.42%	7.35%	8.46%	8.11%
30-Year	8.97%	8.95%	9.06%	9.23%	8.09%	7.80%	7.49%	7.81%

The above are 8 rolling period returns for VFINX and the 7 American Funds in this test. American Funds beat VFINX in 33 of the 63 periods shown.  American Funds beat the index 52% of the time.  Three funds, AGTHX, ANCFX and ANEFX beat 89% of the time periods tested.

CALENDAR YEAR RETURNS

My data spans 31 complete calendar years, 1986 to 2016, and 3Q2017 year-to-date. American Funds overall beat VFINX 40% of the calendar years.  The individual fund beats vary from a spectacular 53% for ANCFX to 44%-34% for the others.

RISK REWARD

3, 5 and 10 year standard deviations of annualized returns have the following results:

3-Year Standard Deviation: 2 of 7 American Funds beat VFINX
5-Year Standard Deviation: 1 of 7 American Fund beats VFINX
10-Year Standard Deviation: 3 of 7 American Funds beat VFINX

CONCLUSION

American Funds NAV generally, over shorter time frames, meet and one beats the returns of VFINX with slightly higher risk. American Funds bold claim is fair.

Disclaimer: This blog is not investment advice, is for information purposes only, may be subject to change without notice, and, while prepared with care, may be subject to omissions and errors.

VTSAX v VFIAX

Here we go again, this time comparing two Vanguard index mutual funds, the Admiral Class Total Stock Market Fund, VTSAX versus Vanguard's Admiral Class S&P 500 Index Fund, VFIAX.

If you ever studied finance in B-school, or anywhere for that matter, portfolio theory was all about stocks versus bonds versus cash. Well, before the Vanguard Total Stock Market Investor Fund was created on 4/27/1992, nobody could ever buy the entire stock market asset class. Since then, you can. Same with bonds and you always had cash.

Both funds come in Admiral and Investor Class shares as well as institutional classes, as do other Vanguard index funds. We are going to compare the Admiral classes this time. But, fyi, key differences are: Admiral shares have a later inception date of 11/13/2000, a $10,000 minimum investment and 0.04% expense ratio versus much earlier inception dates, $3,000 minimum and 0.15% expense ratio for Investor shares.

Since they are both Vanguard funds we can get the numbers directly from Vanguard. Here they are:

Annualized Returns (reported by Vanguard as of 9/30/2017)

                             VTSAX       VFIAX
Quarter-end             4.54%         4.48%
Year-to-date           13.95%       14.20%
1-Year                    18.63%       18.57%
3-Year                    10.69%       10.78%
5-Year                    14.18%       14.18%
10-Year                    7.69%         7.43%
Since Inception        6.41%         5.81%

Dividend Yield         1.84%         1.93%

3-Year Std. Dev.     10.32%       10.13%
Sharpe Ratio               .84              .90

Top 10 Stocks are the same for both funds.
Total # Stocks          3607             512
Top 10 Stocks*          17%           20.5%
AUM**                   $174.5B     $212.6B

I computed the correlation of daily returns since inception to be an indistinguishable 99.6%.

Ok, so whats the difference between these two funds?

-Not much, with little that matters to investors.
-VTSAX includes the small stocks not in the S&P 500.
-Small stocks are more volatile than large stocks.
-Small stocks have lower dividends than large stocks.
-Small stocks have sometimes higher returns than large stocks.
-All three effects account for the slight differences in fund performance.
-The Total Stock Market Index is arguably more academically correct in that portfolio theory likes to own the whole market and not pick winners and losers.

* Top 10 Stocks as percentage of AUM
**AUM = Assets Under Management

Disclaimer: This blog is not investment advice, is for information purposes only, may be subject to change without notice, and while prepared with care, may be subject to omissions and errors.

VFINX Data Test: Yahoo vs Vanguard

Before going any further with index to fund comparisons, we must know just how good IS Yahoo historical price data?

Lets set the scene:

-The idea is to compare historical VFINX data from Yahoo and Vanguard.
-Return calculations must include effects of distributions (dividends, capital gains) and stock splits.
-Yahoo provides Net Asset Value (NAV) closing prices, dividend data and "adjusted closing" prices, adjusted by split and dividend multipliers. Thank you Yahoo. Without these we could not do valid return calculations.
-Vanguard does NOT provide adjusted closing prices for their mutual funds. Vanguard also only provides ten years of distribution data, not a complete set.
-We are left with closes as the only "raw" head-to-head price comparison of Vanguard and Yahoo.
-Yahoo provides dividends for its entire dataset but Vanguard provides only 10 years of dividends.

Taking the price and dividend download of VFINX from Vanguard and matching it against the Yahoo download, we have the following results for the "closing price" series:

Date of downloads: 10/1/2017
Yahoo database start date: 1/2/1980
Vanguard database start date: 8/31/1976

Number of days in Vanguard data: 10,366
Number of days in Yahoo data: 9,525

Number of date errors: 3
Number of Yahoo price errors: 124 out of 9,525 prices (121 excluding missing dates).

Vanguard database dividends reported: 40
Yahoo database dividends reported: 154
Number of reported date errors: 0
Number of reported dividend errors: 0

Here are ALL the errors found in Yahoo's VFINX closing price data.

Due to rounding, most of these are not even errors!

CONCLUSION:  YAHOO DATA IS RELIABLE!


Notes:
-While the VFINX inception date is 8/31/1976, Yahoo only provides data since 1/2/1980.
-Yahoo data is missing for 10/30,2012, 10/29/2012 and 1/2/2007.
-I did not check Vanguard dates, which appear to be complete.
-Vanguard's website only shows dividends for ten years total. This download has a start date of 12/21/07.
-Yahoo data includes all dividends since 1/2/1980.
-Prices were rounded to two decimal places.

Even after this conclusion we have one more step to go: using the available Vanguard dividend data, compare Yahoo's adjusted close to Vanguard's "adjusted" close (called "reinvest price" by Vanguard) For all available dividend dates posted by Vanguard, Yahoo is the same!

Vanguard fun facts: Vanguard has the following S&P 500 Index Fund classes:

Class, Symbol, Inception date, Expense Ratio, Minimum Investment, Assets Under Management

Investor, VFINX, 8/31/1976, 0.14%, $3000, $27.3 Billion
Admiral, VFIAX, 11/13/2000, 0.04%, $10,000, $212.6 Billion
Institutional, VINIX, 7/7/1997, 0.035%, $5 Million, $134.2 Billion
Institutional Plus, VIIIX, 7/7/1997, 0.02%, $100.0 Million, $93.2 Billion
Institutional Select, VFFSX, 6/24/2016, 0.01%, $5.0 Billion, $29.6 Billion
ETF, VOO, 9/7/2010, 0.04%, 1 share, $71.8 Billion

As expenses differ, each class will have its own price series.

Disclaimer: This blog is not investment advice, is for information purposes only, may be subject to change without notice, and while prepared with care, may be subject to omissions and errors.

ACSI v SPY

This post will compare ACSI to SPY. It's interesting because ACSI has an entirely different methodology using customer satisfaction as a basis for investment. This is a new fund. Yahoo's data begins 11/7/2016.

Here's the chart and return data for the comparison.

ACSI v SPY

As anyone can see, ACSI holds up very well to SPY. Of course, this is a very limited dataset and few conclusions can be drawn from it.

If anyone wants the excel spreadsheet, just ask.

Disclaimer: This blog is not investment advice, is for information purposes only, may be subject to change without notice, and while prepared with care, may be subject to omissions and errors.

The Benchmarks-VFINX SPY VOO

For this post I will look at the history of the VFINX, SPY and compare them both to the Vanguard S&P 500 ETF-symbol VOO.

Comparing VFINX and SPY is a good starting point. We need to see if both benchmarks, which measure the same thing, ARE the same. VFINX and VOO is a valid comp because the VOO is the ETF version of the VFINX. The ETF is just another wrapper for the same content. Investors need to know if wrappers matter.

There ARE other index mutual funds and index ETFs  (such as IVV) but they all started much later than the VFINX and SPY and I am not sure the comps of these others will be useful to investors. Since these all measure the same thing, we can expect them to be the same. Below are the history and and data for the VFINX, SPY and VOO.

Full History

Chart 1. Full History

Table 1. Full History

There ARE differences. For some reason, when compared head to head for full years only, the mutual fund has a higher average annual return and lower standard deviation of annual returns than either ETF. Most interesting may be the VOO data which shows a significantly lower return than its mutual fund cousin.

Does the ETF reduce return when compared to the same thing in a mutual fund? Maybe Vanguard has an answer.  The risk reward chart shows VFINX, the top left data point, outperforming with higher return and lower risk than either the SPY, middle point, and VOO, the far left point.

Chart 2. Risk Reward

Despite the return difference, the correlations are very high, as expected.

Table 2. Correlation

VFINX versus SPY

This is head to head vami chart starting with the start date for SPY, 1/29/1993 = 1000.


Chart 3. VFINX versus SPY

Table 3. VFINX versus SPY

When I first looked at this, I thought there was an error starting with the 3/9/2009 recession low. Here are the peak to trough to current date returns.

Table 4. Peak to Recession Trough to Present

VFINX mutual fund sharply outperformed SPY for these critical periods. I still don't get it, but I cannot find an error.

VFINX versus VOO

Here may be the most remarkable test. VFINX and VOO are the same fund except one is a mutual fund and the other is an ETF as noted before. This is a vami chart showing the difference.

Chart 4. VFINX to VOO

As noted before, either mutual funds outperform ETFs or there is a problem I am missing. I'll ask Vanguard about this. Note: mutual funds are marked to NAV every close while ETFs are freely traded with arbitrage enforcement.

The excel spreadsheet for this post is available to anyone. Just ask.

Disclaimer: This blog is not investment advice, is for information purposes only, may be subject to change without notice, and while prepared with care, may be subject to omissions and errors.

Indexing Genie is Out of the Bottle - The Road to Serfdom

In August of 2016, the brokerage firm Sanford C. Bernstein & Co., LLC. wrote what may be the only serious criticism of passive index investing that I have ever seen.

(https://www.bloomberg.com/news/articles/2016-08-23/bernstein-passive-investing-is-worse-for-society-than-marxism).

The gist of this article is that despite the utter and total complete investment success of indexing as an investment strategy (almost 1/3rd of all aum when compared to every other strategy) indexing is a threat to capitalism. That is, passive investing "impedes the efficient allocation of capital". The article further states that "the social function of active management ... is to direct capital to its most productive end, facilitating sustainable job creation and a rise in the aggregate standard of living". 

Hurrah! The indexing genie is out of the bottle: indexing or passive investing is a SYSTEMIC RISK!

The most basic systemic risk in finance is anything that is not sensitive to price.  A well functioning capitalist system creates buyers when prices are too low and creates sellers when prices are too high. Unfortunately, passive indexers, like myself, do not have market incentives to buy or sell. We just buy and hold.

So, what's wrong with that? Lots is wrong with that. Passive investing is divorced from price signals. Passive investing is actually a parasite on the back of the remaining active investors. Without active investing, there would be no price signals and no markets. 

As long as the great majority of investors are active, passive works. Passive assets under management, aum, is currently estimated to be just under 30% of all aum. If this number were to grow much beyond 50% there will be fewer and fewer investors left to tame the markets and passive investors will be on a new uncharted road. 

Disclaimer: This blog is not investment advice, is for information purposes only, may be subject to change without notice, and while prepared with care, may be subject to omissions and errors.

First Post

Saw my friend Phil Bok's new ACSI fund on LinkedIn. Phil's a very smart guy and I like his fund, but I had to test it and that led to this blog where I will test any fund against indexing.

Criteria I will use on every test are as follows:

1-Easy free public data only.
Too many studies use proprietary, modified data that everyone can't check.  If your study uses these, no go for me.  Fund issuer data and Yahoo data is freely available as of this writing. If you claim that Yahoo's data is bad, well, you know, we all want to see the plan. My answer is that I sometimes test for that and Yahoo data is generally very good. Finally, why Yahoo? Because if Yahoo doesn't list it, is it really widely available for the average investor? Show me a more popular and free data source and I'll use it.

2-Continuously compounded daily returns only.
I like easy arithmetic. I compute the daily return = ln(b/a), where a = Yahoo's "adjusted close" on day 1 and b = the "adjusted close" on day 2. Adjusted close is its own issue and my tests for that are pretty good too.

3-Volatility = the excel population standard deviation function-STDEV.P
Simple and as good as any other measure and does not buy in to the fiction that term rearrangements add information.

4-Correlation = the excel correlation function.
Correlation is important because we don't want to buy the same returns in a different wrapper.

5-Select maximum, calendar year and rolling time periods, only.
Tricky one. Every possible time period will have a different result. A perfect track record would beat in every possible time period-impossible!  So how do you pick a fair time period? You don't pick; you use the same periods for each test and look for consistency.

6-Use VFINX or SPY as the index.
You can buy these; you cannot buy the S&P500 Index.

VOO is a great choice too but is 17 years too late. VFINX is the mother of all indexes but it's a mutual fund not an ETF and ETFs are where it's at in today's markets. So, if you have a mutual fund, VFINX will be the index. If you have an ETF, SPY is the bogey.

That's it. If you have a public track record on Yahoo or somewhere that can beat it, send me the symbol. Maybe there's a million out there that nobody knows about.

Disclaimer: This blog is not investment advice, is for information purposes only, may be subject to change without notice, and while prepared with care, may be subject to omissions and errors.

VistaCTA MarketView Archive 2016

Commodity Market Updates for 2016

MarketView December 2016 - First Solid Gain in 5 Years

MarketView November 2016 - Volatile Commodities End Mixed

MarketView October 2016 - Commodities On Hold Before Election

MarketView September 2016 - Energy, Agriculture & Metals Higher

MarketView August 2016 - Commodities Turn Lower

MarketView July 2016 - Oil Drops A Mixed Basket

MarketView June 2016 - Brexit Drives Gold Higher

MarketView May 2016 - Commodities Mixed Metals Dive

MarketView April 2016 - Commodity Rally Takes Hold

MarketView March 2016 - Commodities Rally

MarketView February 2016 - Metals Rally Edge Commodities Higher

MarketView January 2016 - Energies Weigh On Commodities

Back After 3 Years

I'm back and will occasionally be posting after three years absence.  

To start, lets see how things are going since my last post.  Here are the 15 established, liquid, fair and well-formed commodity futures contracts that compose the investable VistaCTA commodity basket:

The weakest performer, no surprise, is WTI Crude Oil, down 87%.  The strongest performer, big surprise, is frozen concentrated orange juice. OJ is up, yes, but on very bad news of major Florida crop destruction due to disease.  The average commodity is down 24% for the period.  The indexes are a little worse.

Here is the chart for the VistaCTA basket (VBX), the Bloomberg UBS Commodity Index (BCOM) and the S&P GSCI Commodity Index (GSCI) (all excess return indexes) for the same period:

Terminal values and total losses for VBX, BCOM and GSCI are $707 -35%, $641 -44% and $434 -84%, respectively. These have been dark years for commodity indexes. Let’s check out the why and wherefors next time. 

VistaCTA Archive for 2015

MarketView December 2015 - Commodities Down Double Digits 3 Years in a Row

MarketView November 2015 - Another Bad Month Sends Legacies Down 20% YTD

MarketView October 2015 - Month of Failed Rallies

MarketView September 2015 - Mixed Markets Post Loss

MarketView August 2015 - Small Loss Masks Deep Volatility

MarketView July 2015 - All Sectors Down

MarketView June 2015 - USDA Reports Counter Commodity Bears

Time to Really Do Contango Homework

There are recently a boatload of misleading, misinformed and inaccurate (repetitive, I know) posts by unnamed others about doing one's "contango homework", about looking at the shape of the futures curve to somehow make some conclusion about the prospective gains or losses to investors in said "contango" or "backwardated" commodities.

So, ok, let's do the homework.  Below are the results of 33 years rolling positions in the VistaCTA commodity basket of 15 major commodities (crude, heat, natgas, rbob/gasoline; corn, wheat, beans; gold, silver, copper; coffee, sugar, cocoa; cotton and frozen concentrated orange juice).  These are the highly select well-formed futures contracts included in the VistaCTA basket.  The results of rolling these contracts since 1980 are shown below.

Per the VistaCTA rules, the basket is rolled once per year, held for a year, sold and replaced with a new basket.  The sell and replacement is the "roll".  Note rolling does not create a return event, only the liquidation (vb sell) of the existing basket versus the original basis (vb buy) creates return.  The vb's shown are each year's new basis.

yr          vb                 cb       %cb     r

1980 613,794.44 contango 5% -26%

1981 497,085.43 contango 9% -16%

1982 431,779.92 contango 3% 6%

1983 459,955.46 contango 0% -11%

1984 424,249.76 contango 4% -11%

1985 375,251.52 backwardation -1% -10%

1986 348,666.38 contango 3% 2%

1987 367,878.51 contango 3% -2%

1988 349,500.62 backwardation -3% 0%

1989 345,122.92 backwardation -1% 19%

1990 395,668.58 backwardation -4% -14%

1991 343,179.91 contango 1% -8%

1992 321,498.46 contango 2% 0%

1993 336,825.07 contango 4% 16%

1994 406,335.85 contango 4% -7%

1995 368,142.81 backwardation -3% 4%

1996 365,516.05 backwardation -4% 6%

1997 375,650.90 backwardation -3% -15%

1998 338,576.80 contango 6% -4%

1999 325,407.08 backwardation 0% 14%

2000 369,313.10 backwardation -1% -22%

2001 307,072.03 contango 7% 17%

2002 348,316.05 backwardation -3% 11%

2003 376,951.43 backwardation -2% 31%

2004 476,701.93 backwardation -4% 41%

2005 608,361.51 backwardation -10% 10%

2006 715,832.95 contango 7% 14%

2007 805,308.50 backwardation -1% 8%

2008 911,174.50 contango 5% -14%

2009 815,403.10 contango 5% 16%

2010 957,105.35 contango 1% 11%

2011 1,065,868.55 contango 0% 8%

2012 1,130,263.50 backwardation -2% -11%

correl correl (0.20)

yr = roll year

vb = VistaCTA basket value when purchased on the roll date.

cb = contango (%cb>0) or backwardated roll (%cb<0)

%cb = % contango or -%cb for backwardation = (vb new)/(vb old) -1

r = subsequent rate of return of the basket after the roll = (vb sell)/(vb buy) -1

Clearly, there is no relationship between contango/backwardation and futures returns.  How sad that billions apparently are committed to a false concept, a relationship that does not exist.  A recent academic paper demonstrates that "roll yield" doesn't even exist, it is an error in the construction and calculation of legacy commodity indices.  

By the Light of the Silvery Moon

I may be the first to proclaim the current turnaround of commodities but... August's sector action brings up a question.

While the VistaCTA commodity basket is up 4.12% month to date (as of Friday's 8/16/2013 close) this is NOT a broad based rally.  Energy, which was July's top performer, is up only 1.34%, grains are unchanged and agriculture, on average, is up 1.92%.  The gains are concentrated in metals which are up 9.77% in August.  And silver, the weaker sister, dominates, up 18.79%.  Silver, per trader lore, is the fickle metal and wanes and waxes with the moon and is usually NOT the marker for a long-term broad based commodity rally.

Coffee Flash Boom Crash Crash

Thursday and Friday were banner days for ICE (formerly CSC) coffee futures contracts.  Weeks of price action were compressed in minutes.  Here's the background.

CHART COURTESY OF BARCHART.COM

September Coffee has been in a slow downtrend from $1.60 per pound to $1.20 per pound all year long with two short lived boosts: $1.55 to $1.65 in January and $1.35 to $1.50 in May.  Since mid-June, coffee has been trading around $1.18 to $1.25.  The chart shows a blip on Thursday July 18th to $1.34 followed by Friday's close at $1.2270 per pound.

  Date    Open    High     Low    Last  Chge  ----52 Week----
07/19/13   12790   13025   12175   12270  -485
07/18/13   12750   13400   12665   12755   -40  -------High----
07/17/13   12545   12920   12530   12795  +200  07/23/12  19700
07/16/13   12340   12675   12320   12595  +270  --------Low----
07/15/13   11930   12410   11860   12325  +385  06/20/13  11710

DATA COURTESY OF BARCHART.COM

The dailys do not show how fast and fleeting the trade was on those days. To get a clearer idea of the extent of Thursday's "flash boom crash" and Friday's crash we need to look at the intra-day charts:

ALL CHARTS COURTESY OF BARCHART.COM

Oddly enough while the above Barcharts do not show it, $1.34 did print and it was fleeting. The boom started with the decisive trade above $1.30 at 7:21 AM (US Central Time) and then the $1.34 touch at 7:27 AM.  Only limit orders need apply and not the $1.34 limit!  The first crash was the 400 point move to $1.27 at 12:10 PM and the final crash was the second 400 point decline to the $1.2275 per pound level near Friday's 1:00 PM close.

The dollar value of the 37,500 pound coffee futures contract at Friday's closing price of $1.2275 per pound was $46,031.25.  The exchange minimum margin for coffee as of Friday was (and still is) $2,750 per contract or 6% of the contract value.  A 7 cent move in coffee (the daily range on Thursday and Friday) equals $2,625 or roughly one's entire margin requirement.  Woe to the speculator trading coffee on exchange minimum margins.

Is the Commodity Turnaround Coming?

The monthly value of the VistaCTA basket of 15 commodity futures contracts since 2009 is shown below:

The gaps in the graph represent the annual roll of the current-month basket to the back-month basket, as defined in the VistaCTA rules.  Note the basket rolled higher in 2010, nearly flat in 2011 and slightly lower in 2012.  Some would call 2010 and 2011 "contango" type rolls and 2012 a "backwardation" type roll.  Note also that subsequent price action appears to have nothing to do with the type of roll.

From a start of under $900,000, the value topped out above $1.3MM and now is valued near $960,000. The July 2013 bounce may be sign of better days to come!

Sixth Worst Day Since 2009

Commodities fell 3.9% yesterday.  Lead by precious metals, then everything else; here's what happened to the 15 commodities in the VistaCTA basket:

silver	-0.087
gold	-0.066
coffee	-0.049
rbob	-0.038
crude	-0.037
sugar	-0.035
heat	-0.032
cocoa	-0.027
copper	-0.026
soybeans	-0.02
corn	-0.018
oj	-0.014
natgas	-0.014
wheat	-0.01

This was the sixth worst day of the 1044 trading days since Vista started 5/1/09.

	top 10 down days			top 10 up days
9/22/2011	(0.0535)		6/29/2012	0.0443
5/5/2011	(0.0517)		7/30/2009	0.0330
9/23/2011	(0.0516)		9/27/2011	0.0300
12/14/2011	(0.0412)		9/30/2009	0.0297
4/15/2013	(0.0411)		10/27/2011	0.0283
6/20/2013	(0.0398)		11/16/2009	0.0281
3/15/2011	(0.0374)		8/3/2009	0.0279
11/12/2010	(0.0365)		1/3/2012	0.0276
5/11/2011	(0.0351)		6/1/2009	0.0266
8/4/2011	(0.0316)		7/3/2012	0.0264

Here is the scatterplot of all the daily returns since 4-09. Note all returns are continously compounded.

As you can see, fall 2011 had the worst declines with the Q3 swoon.  The biggest up day was last summers grain drought.  The two recent low points were the first gold break of 4/15 and yesterday's gold break.

All in all, though, the VistaCTA basket has held very well through this period since inception.

Still 40% up!

There may be two major points of view regarding the current market break:
-One says we are in a continuing secular commodity price decline, i.e. more declines and more volatility are on the way.
-A second view is that, no, this is a man-made, or more accurately, a Fed made crisis that can easily be fixed by man, or, the Fed.   Thus, the better bet is to use breaks to build positions.

The relatively benign response of most commodities to the precious metal decline (for example, energy is still up on the month and ags are down 2%, or so)-the benign commodity response- may be an excellent setup for better returns to come. Statistically, rare events are rare and more normal times prevail.

The tiny red bar (do you see it?) shows yesterdays move.  It's pretty rare.

Natural Gas Futures versus Natural Gas ETFs

There is an ongoing debate about the merits of owning "risky and complicated" futures contracts as compared to owning simple and convenient exchange traded funds (ETFs) when you are trying to gain exposure to commodities (or any asset category, for that matter). While I may be the first to trumpet the benefits of stock index ETFs over stock portfolios, I am not so sure about commodity ETFs replacing commodity portfolios.

The poster boy for bad commodity ETFs has been none other than those created for natural gas (NG).  Now, lets be clear, natgas is way down no matter how you look at it.   The glut of NG has contributed to the American miracle of abundant energy supplies and is reflected in NG prices regardless of the source. Given that, what is the difference to an investor in the major natgas ETFs versus actually owning natgas futures contracts?  Let's look at the charts.

The charts below show the monthly returns of the two major natural gas ETFs versus the investible natural gas futures contracts held by VistaCTA.  The top chart shows the return of $1000 invested in the United States Commodity Funds Natural Gas ETF, symbol UNG.  Not to beat up on any fund family, since this ETF was issued there have been a number of controversies that actually, I feel, have little merit and little effect on investor returns.  As anyone can see, both the NG futures and UNG have declined sharply since UNG began trading.

Monthly Returns of A Hypothetical $1000 Investment

As bad as NG futures have been, UNG has been worse.  

Without getting into any technical details, after a few years of trading UNG, the issuer created a new ETF, symbol UNL, designed to be an improvement/alternative to UNG.  Below is the $1000 return of UNL compared to UNG and VistaCTA's NG futures contracts. 

Well, yes, UNL was an improvement, but still worse than NG futures until recent months.

Based on investor returns of the major natgas ETFs versus NG futures, I think one would be hard pressed to justify the comparable losses for an imagined ease or convenience of investment.  I'll compare investing in futures versus an ETF in future posts.

The Crude Oil Market, It's Just Your Point of View!

Quick observation of the confounding crude oil market:  Is it bullish or bearish, volatile or flat, tradeable or untradeable? Well it all may just depend upon your point of view.

Here is a chart that almost looks like a random number generator:

Here's a totally different view of the same thing:

CHARTS COURTESY BARCHART.COM

Flat sideways or volatile crazy?  You decide.

Soybean Technicals versus Fundamentals

Most recent fundamentals-midwest rains-are bearish while July soybeans are trying to fill last week's $5.50 overhead gap.

COURTESY BARCHART.COM

This is a classic bullish technical versus bearish fundamental setup.  Resolution of this conflict may be the short term completion of the bullish fill pattern and a subsequent long-term decline as the market reconciles with bearish reality.

"Flash Boom" in Soybeans

Thursday, May 23rd was an extraordinary day in soybean futures.  See the July Soybean futures charts below:

Date	Open	High	Low	Last	Chge
5/24/2013	14900	15000	14720	14762	-232
5/23/2013	14930	15410	14880	14994	52
5/22/2013	14750	14944	14710	14942	160
5/21/2013	14630	14790	14564	14782	136
5/20/2013	14490	14650	14454	14644	160

Charts and Data Courtesy of Barchart.com

On apparently wild and unsubstantiated rumors (see Chris Lehner's excellent commentary on InsideFutures.com) beans had a "flash boom" (the reverse of a "flash crash") where a market trades extremely high for a short period and immediately returns to its prior level.  

The 10-minute chart, below, offers a sense of how beans traded in minutes from the high $14s to $15.41 a bushel and back.  $15.40 highs have not been seen since the drought scorched market of last September.    

In the span of minutes, soybeans gap sharply higher to a peak of $15.41 per bushel. Subsequently beans give up all their gains to close nearly unchanged and, the next day, move lower again.  

I have no first hand knowledge but I believe that computerized newsreader algos were behind or significantly contributed to the day's "flash boom".  High freq algo traders trade the story and ask questions later.  Most human traders would try and find facts and then trade or, at worst, the rumor trade would not have been so extreme.

Who gets hurt in a "flash boom'"? The same people who get hurt in a flash crash! Stops are unnecessarily run, margin calls are unnecessarily met, cash is drained from the market. The market structure and bona fide hedgers and smaller specs are damaged-all for no reason.

While good for exchanges, hedge funds, high freqs and algos, flash crashes and flash booms may only hurt the proper function of the commodities market.  Open outcry and limited trading hours have worked for commercials and speculators for over 100 years.  It may be time to revisit aspects of today's electronic trading.  

Was Gold a "Flash Crash"?

The progression of gold's collapse from Friday April 12th to Tuesday April 16th may hold a lesson we could learn from the past.  First, here is how the market looked on those days:



Courtesy of Barchart.com

Here are the open-high-low-closes:

open high low close gap change range 3/28/2013 1605.6 1608.3 1594.3 1595.7 4/1/2013 1598.1 1601.6 1595.2 1600.9 2.4 5.2 6.4 4/2/2013 1600.1 1604.3 1574 1575.9 -0.8 -25 30.3 4/3/2013 1576.4 1577.3 1549.7 1553.5 0.5 -22.4 27.6 4/4/2013 1557.7 1559.3 1539.4 1552.4 4.2 -1.1 19.9 4/5/2013 1554 1581.8 1549 1575.9 1.6 23.5 32.8 4/8/2013 1580.8 1582.9 1566.6 1572.5 4.9 -3.4 16.3 4/9/2013 1572.4 1590.1 1570 1586.7 -0.1 14.2 20.1 4/10/2013 1585 1588.5 1555.3 1558.8 -1.7 -27.9 33.2 4/11/2013 1559 1568.1 1553 1564.9 0.2 6.1 15.1 4/12/2013 1560.3 1564.2 1476 1501.4 -4.6 -63.5 88.2 4/15/2013 1481 1495 1335.1 1361.1 -20.4 -140.3 159.9 4/16/2013 1355 1404.2 1321.5 1387.4 -6.1 26.3 82.7 4/17/2013 1371.4 1395.2 1365 1382.7 -16 -4.7 30.2 4/18/2013 1376.1 1402 1335.6 1392.5 -6.6 9.8 66.4 4/19/2013 1393 1424.7 1385.4 1395.6 0.5 3.1 39.3 4/22/2013 1408.3 1438.8 1403.5 1421.2 12.7 25.6 35.3 4/23/2013 1425.4 1432.8 1404 1408.8 4.2 -12.4 28.8 4/24/2013 1412.5 1433.6 1411.5 1423.7 3.7 14.9 22.1 4/25/2013 1430.3 1468.6 1426.3 1462 6.6 38.3 42.3 4/26/2013 1467.3 1484.8 1447.3 1453.6 5.3 -8.4 37.5 4/29/2013 1466.2 1478.3 1461.6 1467.4 12.6 13.8 16.7 4/30/2013 1476.6 1479.5 1460.5 1472.1 9.2 4.7 19 5/1/2013 1475.6 1477.4 1439.7 1446.2 3.5 -25.9 37.7 5/2/2013 1457 1473.3 1448.1 1467.6 10.8 21.4 25.2 5/3/2013 1466.2 1487.2 1455.4 1464.2 -1.4 -3.4 31.8 5/6/2013 1470 1478.4 1463.8 1468 5.8 3.8 14.6 5/7/2013 1469.1 1470 1440.4 1448.8 1.1 -19.2 29.6 5/8/2013 1451.6 1475.8 1446.7 1473.7 2.8 24.9 29.1 5/9/2013 1473 1476 1452.1 1468.6 -0.7 -5.1 23.9 5/10/2013 1457.5 1461.2 1418.5 1436.6 -11.1 -32 42.7 5/13/2013 1447.7 1448.3 1424.7 1434.3 11.1 -2.3 23.6 5/14/2013 1429.5 1444.9 1419.7 1424.5 -4.8 -9.8 25.2 5/15/2013 1424.5 1429.4 1386.4 1396.2 0 -28.3 43 5/16/2013 1392.5 1397 1368 1386.9 -3.7 -9.3 29 5/17/2013 1385.2 1391.3 1353.6 1364.7 -1.7 -22.2 37.7

After $1650 support was broken in February, $1550 was gold's major support level until Friday, April 12th. The above shows a clear-cut, yet dismal picture of gold trading down a record $140.30 in one day.  Firstly, with a $20 down gap opening and then a $159.90 down trading range. As bad that looks the chart hides the true extent of the problem.

The posted April 12th close of  $1501.40 was determined during the "daily settlement time range" of 1:29 PM to 1:30 PM Friday afternoon.  The gold futures market then continues trading until it "closes" for 45 minutes at 5:15 PM. Since it's Friday, it does not re-open 45 minutes later at 6 PM.  It stays closed until it opens early Sunday evening at 6:00 PM (all NY Time).

Thus, the market on Friday April 12th continued down after the $1501.40 settle (see the hourly prices below) to end Friday 5:15 PM at $1476.10!  Friday's last trade was $25.30 below Friday's posted close!  It pays sometimes to be obsessive about watching quotes.

Hourly Prices

	open	high	low	close	gap	change	range
4/12-1p	1505.5	1505.7	1498.8	1502.9	0	-2.6	6.9
4/12-2p	1502.8	1503.8	1499.2	1499.6	-0.1	-3.3	4.6
4/12-3p	1499.6	1499.9	1482.3	1487.5	0	-12.1	17.6
4/12-4p	1487.6	1488.8	1480	1482.8	0.1	-4.7	8.8
4/12-5p	1482.7	1482.7	1476	1476.1	-0.1	-6.7	6.7
4/14-6p	1481	1484.4	1477	1482.9	4.9	6.8	7.4
4/14-7p	1482.9	1490	1482.5	1489.4	0	6.5	7.5
4/14-8p	1489.4	1495	1483.5	1484.9	0	-4.5	11.5
4/14-9p	1484.9	1485.1	1443.4	1446.1	0	-38.8	41.7
4/14-10p	1445.8	1449.3	1422.2	1441.2	-0.3	-4.9	27.1
4/14-11p	1441	1452	1433.8	1440.9	-0.2	-0.3	18.2
4/15-12a	1441.1	1448.9	1436	1446.9	0.2	6	12.9
4/15-1a	1446.6	1453.8	1441.2	1450.9	-0.3	4	12.6
4/15-2a	1450.8	1458.5	1447.8	1452.5	-0.1	1.6	10.7
4/15-3a	1452.4	1455.4	1433	1437.2	-0.1	-15.3	22.4
4/15-4a	1437.3	1442.5	1401	1411.1	0.1	-26.1	41.5
4/15-5a	1411	1424.4	1390	1393.8	-0.1	-17.3	34.4
4/15-6a	1393.6	1422.1	1385	1408.8	-0.2	15	37.1
4/15-7a	1408.8	1417	1398.9	1409.5	0	0.7	18.1
4/15-8a	1409.5	1428	1403.2	1421.7	0	12.2	24.8
4/15-9a	1421.7	1422.8	1401.1	1404.1	0	-17.6	21.7
4/15-10a	1404.2	1405	1356.6	1364.5	0.1	-39.6	48.4
4/15-11a	1364.5	1379.9	1355.3	1379.2	0	14.7	24.6
4/15-12p	1378.9	1383.4	1368	1371.1	-0.3	-8.1	15.4
4/15-1p	1371.3	1377.3	1360.5	1369.4	0.2	-1.7	16.8
4/15-2p	1369.2	1371.3	1348.5	1350.1	-0.2	-19.3	22.8
4/15-3p	1350.1	1368.7	1350	1357.6	0	7.5	18.7
4/15-4p	1357.8	1359.9	1335.1	1347.9	0.2	-9.7	24.8
4/15-5p	1347.8	1352.6	1347.4	1352.3	-0.1	4.4	5.2
4/15-6p	1355	1367.7	1347.3	1351.1	2.7	-1.2	20.4
4/15-7p	1350.6	1354.9	1350.6	1351.6	-0.5	0.5	4.3
4/15-8p	1351.9	1361.3	1321.5	1331.3	0.3	-20.3	39.8
4/15-9p	1331.4	1341.5	1326	1331.2	0.1	-0.1	15.5
4/15-10p	1331.1	1358	1327.4	1347	-0.1	15.8	30.6
4/15-11p	1346.8	1353.9	1340.1	1348.4	-0.2	1.4	13.8
4/16-12a	1348.5	1367	1348.5	1365.9	0.1	17.5	18.5
4/16-1a	1366.3	1368.3	1357.1	1364.5	0.4	-1.4	11.2
4/16-2a	1364.5	1380.4	1363.8	1376.5	0	12	16.6
4/16-3a	1376.4	1380	1371.1	1372.9	-0.1	-3.6	8.9
4/16-4a	1373	1378	1360.5	1376.4	0.1	3.5	17.5
4/16-5a	1376.4	1392	1375.8	1389.7	0	13.3	16.2
4/16-6a	1389.7	1394.3	1382.8	1383	0	-6.7	11.5
4/16-7a	1383.1	1391	1381.1	1386.1	0.1	3.1	9.9
4/16-8a	1386.2	1404.2	1386.1	1397.1	0.1	11	18.1
4/16-9a	1397.2	1398.7	1379.6	1381.1	0.1	-16	19.1
4/16-10a	1381.1	1389.7	1371.7	1389.5	0	8.4	18
4/16-11a	1389.4	1393.8	1385.8	1389.6	-0.1	0.1	8
4/16-12p	1389.5	1389.9	1373.7	1379.3	-0.1	-10.3	16.2

The greatest damage was done during the NY sleeping hours.  Around 4:00 AM June Gold traded at $1420. In the 5 AM hour June Gold broke below $1400 for the first time.  This is now $100 below Friday's posted close.  We all know the subsequent carnage, notably again in the NY after-hours, with the market bottoming at $1321.50 between 8 and 9 PM Monday night.  

Flash Crash?

To the question at the top, was this gold's "flash crash"? Sunday night may say yes.  Monday night may say no.  

It takes time for people to see, be told or realize what's happening in any market and gold may be even "slower" than other markets.  People who buy gold may tend to look longer term and have stronger views than most other "fickle" traders.  If participants could see where prices were falling Sunday evening (or even Friday's electronic close) they may have been there to rush in and bid the market.  As it turned out, for much of the balance of April and until last week, they did just that and the gold market stabilized (as evidenced by the normal daily changes, ranges and gaps). If there was a "circuit breaker", a "cooling off period", maybe buyers could've  been found and maybe we would have avoided the $1300 handle from the get go.

On Monday evening, though, after a day to consider prices, gold saw fit to make a new low (notably in the after-market) and only then it recovered.  Also this past weeks action, trading below $1400, may also bring this premise into question.

Gold Limits

On June 5th 2006, the COMEX lifted daily trading limits on gold futures.  Prior to that, the daily gold limit was $75 an ounce and gold futures were trading around $660 an ounce, depending upon contract month.. The limit was roughly 11% of the gold price.  

Using 11% or a $150 gold limit today would not have helped investors on April 15th.    A $75 limit maybe would have helped.  It could have maybe produced buyers if trading was halted at $1425 for a day.  

Do limits damage the free trading of the highly levered futures markets? I think maybe so for computers or algos, but maybe NOT for people, that is regular long and short term investors.  A daily $75 limit may be the right circuit breaker for the gold market.  Admittedly that's roughly 5% or so of today's gold price, but that may be enough.  Daily limits have been part of futures trading for over 100 years, they have a seasoned and strong history of bringing order to extreme markets. 

Did ending limit moves hurt or help gold?  A quick look at the 10 year dispersion chart below may provide evidence for an answer:

To my eyes, gold seems more volatile after June 6, 2006 than it was before limits were lifted.  Also, the number of days the market exceeded a 5% daily change can be counted by hand.  April 15th's -10.39% decline was the worst one-day move in the history of the gold market! 

Flash crash or not and despite hedge funds, high freqs and algos, it may be time to restore the $75 daily limit on gold futures.