What’s Dragging Down Core Retail Sales?

One of the biggest questions in corporate earnings is how the US consumers are doing. We know housing continues to recover, if in a sluggish fashion. We know auto sales are doing well, driven by light trucks. How about retail sales? 

Headline retail sales have continued to increase but showing signs of deceleration. This number is heavily influenced by auto & auto parts though –which we already know is strong. On the other hand, gasoline prices have dropped significantly and that drags down the headline numbers even though it should benefit consumers.

So I consider “core” retail sales – excluding auto and gasoline sales – as a more accurate reading for U.S consumers. The growth in this number is shown below.





Notice that post 2008, growth in “core” retail sales have hovered around 3-5%, slower than the 5-7% range pre-crisis. So I dug a little deeper to see which sector is dragging this down.

The Census Bureau report divides retail sales into 13 categories - motor vehicles, furniture, gasoline, building materials, and so on. Going through the data, I found the biggest drags come from 2 sectors: general merchandise stores and building materials.

The contributions of these 2 sectors to core retail sales growth are shown below.

The declining growth in general merchandise stores might have something to do with offsetting growth in online shopping (although apparently not enough to offset overall sales since the headline number does include e-commerce.) .

The other major drag on growth is the “building materials & garden equipment & supplies dealers” category. Growth has recovered but has not yet reached pre-crisis levels. The good news here is that new home sales – as shown by the link in the first paragraph – should have plenty of room to grow.


Looking out the next few years, I can see a housing pick up driving core retail sales up toward pre-crisis levels. Gasoline prices will also stabilize and be less of a drag on headline growth.

Which Securities are Most Likely to Trend?

The title really should be “finding the least efficient markets – Part I”. But that’s such a broad subject I will just focus on one little part of that here.

Greenblatt’s book “You Can Be A Stock Market Genius” outlined some intuitive ways to find inefficient markets – spinoffs, M&A, bankruptcies…etc. But one of the most common forms of inefficiency is right there on the stock chart – the trend.

If you visualize the price chart of a perfectly efficient security, what would that look like? I imagine it would have sudden gaps up or down as new information comes out, followed by flat lines in times of no news. This is because in a perfectly efficient market, rational investors absorb and digest the same information instantaneously, and that should immediately be reflected in prices.

But often we observe security prices that trend. Prices go up for 3 days in a row, a week in a row…and so on. To me that is proof that the market is not perfectly efficient - there are delays in information dissemination, interpretation, and actions on the parts of investors. Whatever the causes, the trend presents good trading opportunities.

The trend is your friend. But how do you identify securities that are most likely to trend? We need ways to quantify “trendiness”. Here is one simple way to do it: count the number of times where prices move in the same direction (“sequence”), divide by the number of times when prices reverse (“reversals”). This is called the “Cowles-Jones ratio” (CJ ratio).

For example if you have price time series data that goes like this: 1, 2, 1, 2, 1, 2. That’s 5 reversals and not a single “sequence”. The CJ ratio would be 0. On the other hand, if your data is this: “21, 22, 23, 24, 19, 18” That’s 4 times where price moved in the same direction and 1 reversal. (22, 23, 24 all moved in the same direction, then a reversal on 19, and finally 18 moved in the same direction as the last number). In the latter case the CJ ratio would be 4 / 1 = 4. So if you go long the security whenever price first ticks upward, your chance of winning is 4 times that of losing.

Calculating this number for SP 500 components from 1/1/2013 to 9/30/2015, I find the average CJ ratio to be 97%. I expected a number close to 1 so this is reasonable. Below are the stocks that with CJ ratios that are 2 standard deviation above the mean – i.e the trendiest stocks since 1/1/2013.

So the “trendiest” stocks have CJ ratio around 1.1 – 1.2 range. A simple strategy would go something like this: go long whenever you see prices shift directions and go up; and short if prices reverse and go down. Your win percentage would be better than 50/50.


Now check out the common currency pairs. Total trading days are more than those for stocks because the stock market get various holidays off.



Note that even the least trendy FX pair is more likely to trend than the trendiest of stocks! That makes sense to me. The forex markets are full of non-economic players like central banks and commercials for whom profit maximization is not the top priority. Then you also have mom and pop participants. When my dad wants to buy some NZD he literally goes to the local banking branch and buy them! That surely creates lags and opportunities not seen in the stock market.

These numbers change depending on what time period you use. But in general I do find currencies to be trendier than stocks.

There are other ways to measure trendiness – perhaps one can quantify autocorrelations, or run backtests using simple moving average crossover rules and then rank the results. As I learn more ways to detect trends (and get more mathematically skilled) I will post my discoveries.

Notes on Archer-Daniels-Midland (ADM)

ADM is one of the largest grain processors in the world. The 4 major segments are 1) Agricultural Services - which store/transport/trade commodities, 2) Corn Processing – this turns corn into sweetners and starches. Importantly this segment also includes ADM’s ethanol operation. 3) Oilseeds Processing – where ADM does crushing & origination. Soybean related products are key here. 4) Wild Flavors – a new segment that does specialty food ingredients.

Why am I looking at this in the first place? My bull hypothesis is as follows: 1) As a processor that takes soybean and corn as inputs, weakening commodity prices should help ADM’s margins. 2) Despite the cyclicality, over the long run demand for ADM’s products should be very stable. 3) As the world consumes more proteins, ADM’s volume should grow at higher than GDP.

I also think ADM is a good company to track due to availability of related data. I can monitor data from commodity futures and USDA to constantly update, prove, or disprove the above hypothesis. Things like soybean crush spread, ethanol prices, currencies are available in real time and inform one’s view on ADM.

Segment Contribution and Drivers

Here is a quick rundown of each segment’s contribution to operating income (last twelve months), as well as their drivers:

Oilseeds processing (42% of operating income)

• The key grain here is soybeans so I compared current soybean crush spread against historical levels. Current industry crush spreads (see appendix at the end of the article) are near all-time high but declining, which points to downside for ADM. My first bull hypothesis – that declining commodity prices should help ADM’s margins –thus appears already played out and in fact fading. 
• However, ADM’s oilseeds margins (in terms of $ per ton processed) were stable the past few years even as industry crush spreads fluctuated. My guess is ADM will be relatively stable when industry spreads decline as well. But I certainly would not count on much upside from margin expansion. 

Corn processing (24% of operating income)

• Operating profit from ethanol was down ~$220mm in 1H2015. Some of that will be recoverable as ethanol margins recover, creating upside in the segment.
• However, it seems that analysts are already building in ethanol price recovery (perhaps due to expectations of oil price recovery). Consensus has EPS of 3.08 and 3.40 for 2015E and 2016E. This 10% growth will be hard to come by without ethanol at least stabilizing. Besides, the days of $100 oil are over and ethanol margins are unlikely to fully recover to past levels.

Agricultural services (27% of operating income)

• This is storage/transportation/trading…etc. Historically profit here relies on level of US exports, which is a function of how competitive US is versus say, Brazil. With USD strong and the Brazilian real weak, this segment is unlikely to see much advantage. The near upside here is 1) El Nino somehow destroys Brazilian crops, and 2) US has a great crop. 

Wild Flavors (6% of operating income). This could be a growth area for years to come but not enough to offset the importance of the other segments.

Upside Assessment and Current Action Plan

Looking through the main segments, I just don’t see much earning upside beyond those already factored into the consensus. Thus returns will have to come from multiple expansion, which will likely happen if ethanol stabilizes and exports don’t fall apart. In terms of valuation, right now it trades at 11.9x TTM vs 10yr median of 13.3x. That’s a 12% upside from recovery in multiples. In terms of downside, soybean crush margin can weaken, ethanol stabilization can come later than expected, and Brazil and China can enact policies that hurt US competitiveness and export levels.

This is not enough for me to get in. For now I will sit on the sidelines and wait for fundamentals to get better.

Appendix: Historical soybean margins

I calculated these from front month futures data and the formula Soybean crush = soybean oil (in cents/lb) * 0.11 + soybean meal (in $/short ton) * 0.022 - soybean prices (in $/bushel).  The crush spread is still higher than historical because soybean meal is a larger contributor than soybean oil, and soybean meal prices has not came down as much as the others.

Australia’s Trade with China: Export Iron Ore, Import Women

Australia’s latest balance of payment data included a special case study on trade with China. Scroll to the bottom of that link and you’ll see some interesting stats. 

These male to female ratio looked a little skewed? Especially we’re talking about China here - a country known for too many men.

And Australia gets them young too.

I knew Australia has a trade surplus with China, but I didn’t know it was this good!

Miscellaneous Thoughts about the Market

Here are some thoughts I had about the market in recent weeks. The first 2 sections explain why I think the S&P 500 is still overvalued eve after the recent drawdown.

Earning growth expectations for 2016E still unrealistic

Below are revenue and earnings performance of S&P500 for past 10 years, 2016E data are based on analyst estimates. (I got the data from here)

Notice that in 2016E S&P revenues are supposed to grow 6.3% and earnings 10.5%. Digging deeper into the Factset file referenced above, you can see this is because the analysts are assuming 1) energy sector rebounding sharply, 2) all time high margins.

These assumptions strike me as unrealistic. Regarding the former, the energy sector is expected to grow 19% in 2016 – and this is after being revised downward from 34% a few months ago. The risk is if oil stays lower for longer that 19% earnings growth may not pan out. Regarding the latter point, how we’re supposed to see all time high margins is beyond me, especially when margins have been trending down in recent quarters.

Even if energy sector rebounds sharply, forward P/E multiples should not be high because that’s a one-off rebound, not a recurring growth pattern.

The idea of “low rates justify higher forward P/E” is double counting

Two most often heard reasons for high valuation multiples the past year are 1) low rates, 2) growth. The theory is that the “justified” P/E ratio is calculated as dividend payout divide by (required return on equity – growth). Where required return on equity is defined as risk free rate + “equity risk premium”. Leave aside the equity risk premium which is not directly observable, low rates and strong growth should lead to higher valuation multiples.

This all seem very sensible but does not jive with the data – at least not the rates part. If you run a regression where y = earning yield, X1= 10yr US Treasury rates, X2 = earnings growth, the result (which surprised me) would show that interest rate is NOT a significant factor at all!

I find that counterintuitive – obviously low rates should force up equity prices, no? I think it’s because much of the impacts of low rates are already reflected in growth. If you think about it, low rates are supposed to drive more investment and consumption, which lead to higher earnings growth. So the idea that lower rates justify higher forward P/E ratios is double counting – because the impact of low rates is already reflected in higher forward earnings.

Random Thoughts

How can stock prices be random? When people say stock prices are random, what does that even mean? S&P500 is around 1920 today. So you mean to tell me SPX is as likely to be at 20,000 tomorrow as it’d be at 1940? That’s nonsense. Or are they saying returns are random? So SPX could return 5% next year, or it could return 1030%? That’s clearly absurd also.

I think what they meant is that returns follow a normal distribution or some other probability distribution. But you don’t know what that distribution is. Now that seems likely. But still, “random” would imply stock prices have 0 correlations with fundamentals such as earnings movements. That’s clearly false.

So what do people really mean by “random”? Most likely, it’s just code for saying “it’s so complicated that I can’t figure it out”. I’m not sure they tried.



You don’t believe government can pick winners and losers. So why do you think you can?  Lots of reasons here. Somebody has to win. Besides, investing is not just picking winners and losers, valuation counts too. You buy a “winner” company at 100x free cash flow and you’re going to be a loser.



Day trading is a sad way to live. With the market turbulence, I found myself looking at charts of S&P futures all day. I drew lines of support and resistance and gaps. I looked for breakouts, failure to breakout, and “fake outs”. At one point I found myself checking the charts every hour, staring at minute by minute candle charts of ES Sep’15 oscillating between 1930 and 1950. This is all quite miserable, so I stopped. No day trading for me.

Learning Some Commodities

I spent the past few weeks learning about the futures market, as well as various commodity classes. The reason is I am having a tough time taking directional views of stocks given the generally high multiples, so I prefer relative value ideas instead. But relative value comes much more naturally in fixed income, currencies and commodities. I also hope what I learn here help in my equities investments when the time is right. Some notes here.

Waiting for Back Ended Oil Contango to Normalize

Exploration and Production (E&P) stocks seem to be still pricing in $60-70 oil in the longer term. However December 17 and Dec 18 futures are trading around $54 and $57 respectively. So I’m still waiting for E&P stocks to fall further. But we’re getting there.

Term structure of futures is supposed to tell you market’s future view of storage capacity, cost, and inventory. These things are in turn dependent on the market’s forward view of production and consumption.

Crude oil is currently in an oversupply position so I’m not surprised that the front end of WTI futures (CL) is showing some contango. But this condition extends to the back-end too. Below shows how the spread between Dec 2018 and Dec 2016 evolved over the past year. (This is Dec 2018 minus Dec 2016, so higher = contango).

I read this as the market saying oil will be way oversupplied even in December 2016, thus demand for storage will still be high (implying the market still expects a turnaround in oil by December 2018). That seems inconsistent. By 2H16, if we still have an oil glut I would think the market would be convinced that cheap oil is here to stay (perhaps due to technological advances lowering marginal cost of a barrel), then the back ended futures should come down as well. Term structure should normalize.

I would prefer betting on term structure to normalize than taking a pure directional view. Maybe short Dec 2018 futures and buy Dec 2016 and just keep them there. This would be better than trying to pick a bottom by going long front end futures (and risk sell low / buy high on each rollover)

Additional Notes

The mathematical definition of contango is “convenience yield” < (interest + storage cost). Where “Convenience Yield” is a plug, a calculated number you back into. As such, “convenience yield” reflects a mishmash of miscellaneous factors including not just the value of having it now (“convenience”), but also sellers’ desire for insurance, arbitrageurs expectations, speculator sentiments…etc. So a contango could be due to a few possibilities, the main ones are:

• Trade buyers: well stocked already so no one wants more oil now. Value of convenience is low.
• Trade sellers: low need for insurance (else they would have sold futures and push back-end future prices down)
• Storage providers. Inventory level is high and storage capacity could be running out. So storage cost is bid up.
• Speculators and arbitrageurs. Their actions could just be trend following or be based on nuanced view of future balance.

Some people (like Gartman here) argue that futures term structure have NOTHING to do with future expectations of prices, but just a function of supply/demand and storage cost. That sounds like a smart thing to say but they are really just playing on semantics. All supply and demand have some element of participants’ future expectation - if oil trades at $10 and you expect spot prices to be $100 a year from now, you would buy now, store some oil and sell in the future. You would not only impact the supply and demand of oil, but also bid up storage cost in the process.

The Great British Pound Appreciation of 1996-1997

Above is the historical trade weighted exchange rate for British Pound. Since the latest 80’s there are 3 episodes of big moves. First a gap down around 1992, then a big spike up from second half of 1996 to early 1998, and finally the collapse from 2H07 to end of 2008. The last one is straightforward– the Great Financial Crisis really started in 2007 after the subprime bubble popped. The first one I know also. That was Soros and Druckenmiller breaking the Bank of England.

But what about that big run up in 1996-1997? What happened there?


The Bank of England does a great job of archiving their old reports, and one can quickly figure this out by searching through their Inflation Reports from 1996-1997. What happened was a rare combination of 1) monetary divergence, 2) fiscal divergence, and 3) political uncertainty in the rest of Europe.

The monetary policy part is fairly similar to what’s happening today. Back in 1996 the market expected other European countries to lower their rates while Bank of England was expected to tighten money.

But that was only part of the story. The real driver there was countries gearing up to join the Euro system. Keep in mind this is 1996 and the Euro did not exist yet. Other European countries were preparing to join the EMU and they had to meet the “Maastricht convergence criteria” – which would require them to reduce budget deficits. Here’s how the old BOE report explained it:

“A reduction in planned fiscal deficits abroad—such as that taking place on the Continent to meet the Maastricht convergence criteria—would tend to reduce interest rates there, both because lower deficits raise national saving and because fiscal consolidation may reduce aggregate demand in the short run. The exchange rates of the countries undertaking fiscal consolidation would depreciate as financial capital sought higher returns elsewhere”

And this: 

“the portfolio shift away from currencies of countries most likely to participate in EMU may also have reflected higher risk premia because of the uncertainties involved”

To prove the importance of the EMU and Maastricht convergence, BOE’s November 1997 report even had an interesting study showing that the Italian Lira and Deutsche Mark were getting increasingly correlated, while Sterling/Mark significantly less so. 

So there, a combination of monetary divergence and expected fiscal divergence (due to the advent of Euro) caused the GBP to revaluate upward in 1996-1997.

Here’s what I couldn’t help wonder though: how much of U.K. not joining the Euro could be attributed to its humiliating experience in 1992 trying to tie a currency? Is it possible that Soros, Druckenmiller, and the currency speculation world “taught” U.K. the importance of having an independent monetary policy? Without Black Wednesday, could U.K. have ditched the GBP for the EUR?

If so, Soros and gang might have "broke" the Bank of England, but they saved the GBP.