Does The Choice of Index Matter?
We often hear shippers argue that they can cut the cost of truckload (TL) fuel surcharges by using a low estimate of diesel prices when calculating the charges. The prices are drawn from an appropriate industry index. It’s an appealing theory – but does it work in practice?
The price of diesel is one of the main components of fuel surcharges. Over the next few weeks we’ll look at each of these variables to clarify how changing them impacts TL transportation costs. We’ll round off the series of posts with a general assessment of where we think these programs are headed.
Choosing the most advantageous fuel index matters, some shippers point out, because every surcharge has a fuel cost level at which the program is neutral; there is neither a charge nor a rebate. This is called the peg, and we’ll explain it in more detail next week. When prices rise above this peg value the shipper pays a fuel surcharge, when below they receive a rebate. It follows that using a fuel price index with lower numbers on average makes it more likely that there will be a zero charge or a rebate, assuming the same peg is used. The lower estimate is nearer to the peg, so the effective increase in fuel price is less.
The standard measure of diesel prices is the DOE (U.S. Department of Energy) National Average index. But proponents of the low-estimate theory maintain that this is not an accurate yardstick because it doesn’t reflect actual prices paid. They say that carriers negotiate deals that allow them to pay less than the average prices listed in the government index.
The flaw in this argument is that it overlooks the bigger TL picture. It assumes that TL carriers don’t change their line haul rates to reflect the change in fuel surcharge. Madhavi Kanteti and Jordan Levine[i] looked at the issue in their MIT Master’s thesis Risk Sharing in Contracts: The Use of Fuel Surcharge Programs. The researchers surveyed various service providers to determine if they treated customers with individual fuel surcharge programs differently.
The short answer is yes. This makes sense in a market that is defined by near perfect competition. For a more detailed explanation of why the TL market can be described in this way see C. H. Robinson’s Explores report for the Council for Supply Chain Management Professionals (Deriving Strategic Advantage from Truckload Procurement). For our purposes here, perfect markets react quickly to known information and do not display wide variances, which is broadly how the truckload business behaves.
Since providers tailor their pricing strategies to each shipper, they tend to compensate for cuts in fuel surcharges by increasing overall line haul rates. In other words, the savings shippers might gain by selecting an index that gives a low fuel price estimate are negated by increased line haul costs.
Another argument used by the low-estimate camp is that truckers buy fuel in specific geographic areas or lanes, which may differ in price from the national average estimated by the DOE index.
But this line of reasoning neglects the fact that oil is a global commodity, and as such, its price fluctuates according to a basket of industry and geopolitical factors. While there are some local variations in fuel prices, in the long run this variance remains stable.
We’ve done some research to explore this situation. Our researchers looked at the National DOE Average going all the way back to March 21st 1994, a total of 952 weekly observations. We reviewed the correlation between the DOE national average and the five different PADD’s (Petroleum Administration for Defense District, a geographic aggregation of the 50 States and the District of Columbia into five Districts, with PADD 1 further split into three sub-districts. For a list of states covered click here).
The researchers found that each of the regional indexes was highly correlated to the National DOE number. Next, we looked at the self-serve index and wholesale index published by T-Chek Systems, a subsidiary of C. H. Robinson. Here we only had 650 observations starting in January of 2012– but again; a high correlation with the DOE National Average emerged.
The results are shown in the graphs below. In the Figure 1, we compare the National DOE average to PADD 5 (the U.S. West Coast). On average, the West Coast PADD has tracked $0.127 higher than the DOE number. So, in Figure 2, we graphed two lines, the National average plus $0.127 and the West Coast PADD. As you can see, the two lines are very similar (for the math geeks among us the r-squared is 99.5%).
Figure 1. Comparison of the National DOE average to PADD 5
Figure 2. Comparison of National Average -$0.67 and the T-Chek Wholesale or Rack Price
Again the lines are very similar with an r-squared of 99.3%. In Figure 3, we also graphed the actual variance for both of these situations. Comparing the West Coast PADD to the National average with the $0.147 offset, we found that 91% of the time both indexes were within plus or minus a single penny. When comparing the National average to rack pricing with the $0.61 offset, we were within the same range 72.9% of time and 92.5% of the time we were within plus or minus two cents; what you would expect in a global market. Regional markets might be higher or lower but that variance does not change significantly over time and for the most part these regional indices move in tandem. Moreover, even changing to a wholesale index doesn’t alter this overall picture.
The bottom line is that when formulating a fuel surcharge program, picking an industry index that gives a low fuel price estimate will probably save you a lot less than you think. Frankly, there are much better ways to reduce costs from of your TL operations.
Next week we’ll shine a light on another surcharge variable.
 Risk Sharing in Contracts: The Use of Fuel Surcharge Programs,” Madhavi Kanteti and Jordan Levine, MIT Supply Chain Management thesis, Class of 2011. To view the abstract and Executive Summary, visit http://ctl.mit.edu/library/risk_sharing_contracts_use_fuel_surcharge_programs. To request a copy of the full thesis, please contact Dr. Bruce Arntzen, Executive Director, MIT Supply Chain Management Program at: firstname.lastname@example.org