Is a Rising Baltic Dry Index a Bullish Sign for the Market? No.

Schaeffer’s Investment Research recently put out this note on using the Baltic Dry Index (BDI) as an indicator for the S&P 500 [emphasis mine]:

The Baltic Dry Index (BDI) continues to rise. Many consider this an excellent sign for the world economy. The BDI is an index that measures the price of shipping dry bulk cargo, which includes coal, iron ore, and grain. Since the index is used to gauge the demand for major raw materials used in the early stages of production, it is regarded as a leading indicator for the economy…

The theory sounds great: a rise in the BDI signals a growing economy and, therefore, is a bullish sign for the market…History seems to back up this theory, as the SPX's returns following this signal outperform the index's typical returns. In the eight instances that the BDI rose above its 200-day moving average, there were positive returns seven times during the six months following the signal. The average six-month return for the SPX following such a move in the BDI is 7.24%. That's more than double the SPX's typical six-month return of 3.02%.

Incomplete analysis
While the analysis based on the recent action of the Baltic Dry Index and the S&P 500 sounds intriguing, the long term picture shows virtually no relationship between the BDI and the stock market.

The chart below shows the indexed returns of the two indices from 1985. The correlation of monthly changes between the BDI and S&P 500 from January 2007 was 0.33, which is relatively high to suggest some causal relationship. Looking longer term, the correlation of monthly changes was a miniscule 0.07, indicating little or no relationship between the two indices.

click to enlarge

Vincent Fernando, who wrote the following note at Research Reloaded, agrees:

I have actually done a rather in-depth BDI correlation study in the past during my tenure as shipping analyst at Citi. One important thing to note about correlation is that you need to look at correlations for many different time periods because if you just calculate a single correlation then it will be highly dependent on your start and end dates…

I think its even best to step back from numbers or historical correlations and think about some very important differences between the nature of the BDI and the nature of a stock price for a company or market. The BDI measures the rate to ship something around the world, ie. the COST to ship something around the world. On the other hand, the stock of a company represents ownership of a productive asset, a company. (well, at least usually they are productive!) A company is seeking ways to be more efficient, grow the size of its business and generally increase its value.

The analysis from Schaeffer’s was at best incomplete and at worse sloppy and not well thought out. This is an issue that I have talked about over and over again: Good quants need to think about the assumptions in their models.

Beware of quant models bearing gifts.

Comments

[onebir:]

"Good quants need to think about the assumptions in their models." And looking for leading indicators with good long-term correlation is equivalent to assuming equity markets focus on the same indicators long-term, and that those indicators contain the same type of information &/ amount of noise long-term.

I started following the baltic dry about six months back, in the hope it would give me a handle on the world economy. Even over such a short period, it's become pretty clear it's driven by disparate factors at different times. Here are some of them:

A) When it hit its trough in late 08, a major factor was the difficulty exporters faced getting trade finance - no trade finance, no trade - a symptom of problems in the financial system.
B) While the baltic dry's not tradeable in the sense of an equity or a futures contract, I've read that its peak was partly driven by speculators chartered ships in the hope of rechartering them for higher rates. That points to a bubble - in hindsight at least.
C) Some recent upward movements in the baltic dry have been ascribed to wheat shortages in China, and later Chinese resource stockpiling. And longer-term, commentators are expecting (/observing?) some contraction of supply, as shippers scrap excess fleet.

In A, baltic dry strength seems (mildly) positive for stock prices, in B it's perhaps negative, and in C, not very relevant; not too surprising it correlates poorly with stock prices. That doesn't automatically make it completely useless; but does suggest it needs to be interpreted carefully. & probably by an expert - which I'm not.

Jun 02 09:26 AM

[Mistrofan:]

Cam - excellent points - I fully agree. Thank you once again.

Jun 02 10:09 AM

[morph366:]

The situation you describe is one of the drawbacks of having a function in Excel that calculates the correlation coefficient to 15 decimal places.
The selection of the time period and dataset is vital as you suggest.

Jun 02 10:12 AM

[Alphameister:]

Thank you for not posting this article three months ago! ;) While the correlation may not be high over all periods, it seems to be high at key turning points. There was a significant rise in the BDI to a multi-year high in early '95 and again in early '03, at the inceptions of major market advances. While there was no plausible way the index would surge to a new high in the current environment, it has made the type of significant upsurge seen in advance of other bull markets. There is more to analysis than blindly running a correlation check.

Jun 02 02:55 PM

[Phil Trupp:]

The BDI has very little correlation to the broader market. It is narrowly defined. The primary focus of the BDI has to do with shipping rates as they impact the financial health of carriers such as Frontline and other dry bulk shipping companies. A secondary focus is on global demand for dry bulk commodities such as copper, coal, steel ingots, etc. It's quite a stretch to equate these factors to the overall market.

Jun 02 03:13 PM

[Alan Young:]

I agree that the BDI is far from an all-purpose indicator. However, the current low point in worldwide markets is attributable to the dropoff in trade. The resurgence of shipping can indeed be an indicator of a move toward normalization.

Actually, the use of a transportation index to predict broad market trends is a major feature of the original Dow Theory, over a century ago.

Jun 02 06:25 PM

[change is t...:]

To analitically correlate the BDI with anything except shippers' rates means your ability to persuade had better be stronger than the data.

Jun 02 09:59 PM

[onebir:]

www.mgn.com/news/daily...

"PRECIOUS SHIPPING EXPECTS DRY RATES CRASH
Monday, 08 June 2009
Thailand-based handy operator Precious Shipping says that the current rise in the Baltic Dry Index is unsustainable in the long run and will lead to a consequent crash in rates as new tonnage enters the market.

"China binge buying iron ore against all expectations with imports rising at 27+ pct higher in 2009 than for 2008, based on the annualized results of the first 4 months of this year, shows you where the big move from the demand side has come from," the company said. "This rise in iron ire imports is despite the fact that steel production in China in the first four months of 2009 has been roughly at the same levels as we had seen in 2008. An explanation for these increased iron ore imports could be the fact that domestically produced iron ore in China is of a rather poor quality and quite expensive when compared to spot imported prices. Another explanation could be that of speculators getting into the import market to try and get hold of 'cheap' iron ire that would possibly be required under the Chinese government's US$586bn stimulus plan. And a third could be the impending conclusion of the iron ore contract price negotiations."

"Obviously, when you binge buy and compress imports of a single commodity, carried mainly by capesize ships, into a very short space in time, you tend to create two issues at the same time. Firstly, you tend to push up freight rates due to the time-compressed/explosive demand growth for that ship-sector. And more significantly, you create queues at loading and discharging ports which tend to reduce availability of spot ships driving prices even higher. Combine this with delayed delivery schedules of dry bulk ships during Q4 08 and Q1 09 and you have the ingredients of a perfect storm especially when you take the number of capesize ships that have been sent to the scrap-yard during the last 6/9 months."

"The BDI hit its recent peak on the 3rd June 2009 closing at 4,291 points with the average t/c rate per day per ship for the Capes at USD 93,197, Panamaxes at 28,110, Supramaxes at 19,470, and Handies at 12,668. Compare these numbers with what was prevailing on the 5th December 2008 when the BDI hit a two and half decade low of 663 points, then Capes were earning a paltry t/c rate of USD 2,763 per day per ship, the Panamaxes were at 4,058, Supramaxes at 5,726, and the Handies at 4,352. Clearly the BDI is more reflective of the Capes and the Panamaxes, as the Panamaxes are a shadow of the Capes, than of the market as a whole."

This, Precious says, will lead to problems. "Besides this sudden, sharp and unexpected rise in the market will have two very bad consequences. Firstly, scrapping will come to a grinding halt. This will stop the squeeze on the supply side which was originating from the permanent removal of ships via the scrap-yards. Secondly, with the new 'confidence' in the market, all the delayed deliveries of new ships from the builders might make a startling reversal and lots of new ships could suddenly start appearing on the market. Both these consequences will result in long term damage to our market and eventual balance between demand and supply will be pushed further away on the time horizon."

"In our opinion, based on the congestion at Chinese discharge ports and the present increased level of iron ore imports the index could actually cross 5,000 points. We think that congestion should clear up once this binge buying of iron ore abates or more new Capes are delivered from the ship-yards than the demand can absorb. Once one or both of these events take place, congestion will vanish very quickly with the BDI crashing down as quickly as it has advanced. The fundamentals still overwhelmingly point to a world economic recession with tremendous job losses and we suspect that will put a real dampener on the current burning hot BDI."

Jun 09 07:19 AM