Tired of Yo-Yo Contracts? Fix the Price with an Indexed-Based Model
When markets yo-yo, so do buyers and sellers ... and they waste time, and money, doing so. When prices fall, buyers scramble to cut new contracts to obtain better prices and mythical "savings"*. When prices rise, suppliers scramble to exit contracts to get better prices from other buyers. And when prices yo-yo, it's a never-ending renegotiation dance that does nothing but waste time, and money, especially when there's an easy way to lock in a contract for the long term that allows both parties to win. It's called the Price Index(ed) Contract, and in this post I'll explain how you can lock in a contract that will protect both parties and allow you to avoid the renegotiation dance ... which will allow you to focus on more categories and new, and better, cost savings opportunities.
The first thing to do is to build a should-cost model that captures all of the major price components (> 5%, if not > 10%) using current prices, drawn from an index. We'll use a (hypothetical but representative) cost breakdown for a 30kVA power transformer, since we all need power, as the foundation for our example.
| Cost | Amount | Percentage |
| Steel | 1055 | 34% |
| Labor&Overhead | 620 | 20% |
| Misc | 560 | 18% |
| Copper | 530 | 17% |
| Oil | 215 | 7% |
| Transportation | 125 | 4% |
| 3105 | 100% |
This tells us that our primary cost components are steel, labor, and copper and that oil would have to to fluctuate 5% to have the same impact as a 2% fluctuation in copper and 10% to have the same impact as a 2% fluctuation in steel. This also says that if our threshold for negotiating a contract is a minimum cost reduction of 5%, that oil would have to fluctuate 70% for us to even consider a renegotiation. Given that rampant runs, like the one which we just experienced where oil tripled and fell back to the baseline in less than two years, historically only happen every few decades, and that labor costs tend to increase rather predictably with inflation, it also tells us that we should only be concerned with steel and copper costs.
There are market based indices for both steel and copper, the CRU is one example of the former and the New York Futures Market is one example of the latter. At least one of them will, on average, correlate to the prices that your supplier consistently plays for steel just like at least one of them will, on average, correlate to the prices that your supplier consistently pays for copper (which depend on their contracts, leverage, and the market(s) they buy from). You just have to agree on one.
Then, you can tie your contracts to the index and word them to automatically adjust prices on a monthly (or quarterly) basis using the impact of market price fluctuations on the should cost model. Since steel accounts for roughly 34% of the cost, if you agree to cost a increase of 1% for every 3% increase in steel cost, you won't have to worry about your supplier having to choose between reneging on your contract or financial ruin in a bull market and if your supplier agrees to a cost decrease of 1% for every 3% reduction in steel cost, he won't have to worry about you having to choose between finding cheaper sources of supply or risking financial ruin due to your inability to compete with (unrealistically) high costs. Similarly, if you agree to a cost increase of 1% for every 6% increase in copper price and your supplier agrees to a cost decrease of 1% for every 6% decrease, neither party loses -- and more importantly, both parties win. In bull markets, your supplier gets to pass on the raw material cost increase -- and only the raw material cost increase (as you both know the cost, no ridiculous mark-ups), and in down markets, you reap the savings without having to go through a long, time-consuming, wasteful renegotiation.
This works for any category you can think of, allowing you to lock in 1, 2, 3, and even 5 year contracts (on non-strategic categories) without having to worry about lost opportunities or overpayments. The only thing you have to do is check the indexes once a month (or quarter) on the date you both agree to "reset" the prices for the next month (or quarter) to make sure your supplier is holding up their end of the bargain. And you can even use this model to take into account financial costs, such as foreign exchange rate assumptions (if raw materials or components are being bought in a foreign currency) or finance assumptions (if part of the product or transportation cost needs to be financed). So build a should cost model and get that yo-yo off your finger.
* There's no such thing as "savings". "Savings" is just money you shouldn't have spent in the first place. (And that is why "cost avoidance" is more important than "cost reduction", despite the refusal of many old-school diehards to recognize that fact.
Categories: Risk Management, contract management, Negotiations
Tags: Indexed Contracts
I agree that indices are valuable and should be incorporated into contracts where a feedstock, labor, energy, etc. comprise a large part of the product cost. In these cases, neither the buyer nor supplier can risk being out of sync with the market. However, your suggestion of three to five year contracts contains an assumption that is not necessarily accurate. Prices are frequently market driven rather than cost driven. A year ago, a 10% profit margin might have been competitive whereas today it may be 5% or less due to the significant supply/demand imbalance. As a buyer, if I'm locked in with a 5 year contract, I'm paying 5% or more than my competition. I can't afford to do that over the long haul. For products that are market driven you need to be out in the market more frequently.
Barb, good point ... but note that I said that "you can even use this model to take into account financial costs".
If you have a market driven category, you can take that into account too if a product price index (PPI) exists that calculates the average increase or decrease in price over the last year, because you can use this index to calculate an average increase or decrease in margin across the board once you factor out the changes due to raw material inputs for all of the significant raw material components.
It's a little bit of calculation up-front, but a valid methodology as long as you account for every significant input (> 5%).
For example, let's say, in our example, using agreed upon indexes, that the average cost of steel increased by 10%, labor by 4%, and copper by 15%, while oil fell 20% and transportation fell 30%. This says that we would expect our product to increase in cost by .10*.34 + .04*.20 + .15*.17 - .20*.07 - .30*.04 = .034 + .008 + .026 - .014 - .012 = 0.042 or 4.2% on the PPI. If, in reality, it only increased by 2.2%, that would tell us that average profit margin decreased by about 2%, and we could calculate that in to our price as well.
So, while a consultant might proclaim that the only way to get the absolute best price is to renegotiate the contract at least yearly, if not every quarter (and hire her again and again and again) and be right, if you're willing to settle for "almost the best price", which will never be off by more than a fraction of a percent iff you adjust for all the significant cost components, you can lock in contracts on many non-strategic categories for up to 5 years without any significant worry (provided that you can get your supplier to agree to a market-driven margin adjustment using third-party PPI indices).
Although my prescriptions might come with a lot of instructions (and caveats) attached, unlike some quacks, this doctor doesn't give bad medicine. :-)
The problem with averages is that you can drown in water that is on average a foot deep! Some suppliers may be hungrier than others which could cause them to offer more aggressive pricing with lower margins. Call me old fashioned, but I'd rather be out in the market more frequentlly (even when I wasn't a consultant) and know that I was getting the best price. This is less true with non-strategic purchases, as you suggest. There is certainly a happy medium between too often and not often enough. The beauty of today's automated eSoourcing tools is that I can conduct a repeat competitive bid without a lot of time and effort. Perhaps we may need to agree to disagree on this.
Barb:
True, but as you pointed out, I'm only advocating longer term contracts for non-strategic purchases where there is low variation in prices outside of the significant cost factors in the should-cost breakdown AND where you have appropriate indexes for each direct and indirect cost that could fluctuate to the point where you want to account for it.
If all those points aren't true, then you shouldn't rely on this method to cut a long term contract.
As for drowing in averages, that only happens in the short term ... maybe you might cut a contract today and see the average PPI drops 10% in 2 months time but after agreeing to a semi-annual margin adjustment ... which would mean you would lose for 4 months ... but since every down in the market is eventually followed by an up (and vice versa) ... it's quite likely that in 22 months, the reverse will happen and prices will jump 15% 2 months after the adjustment, which would mean that you gain for 4 months. The nice thing about averages is that, over the long term, the probability of them NOT averaging out becomes very low.
So if we're talking a non-strategic mid-spend (1-10M) category that historically doesn't have a lot of variability, would you really want to waste tens of thousands of dollars on an sourcing event that would likely result in a time-consuming renegotiation event when the net savings might only be a few percent, or not much more than the cost (when you factor in personnel time, system costs, lost opportunity cost) when you could be attacking a new category?
Especially when your average company has less than a third of its spend under management, which is likely largely due to the fact that they only have time to do so many events in a year? While it's good to be out in the market as often as you can, it's also good to get the majority of your spend under management, and in a shop with limited resources, this is one way to get a reasonable number of categories under long-term management with limited risk AND, at least the first time, significant savings. Permanent solution? No. Good solution? Yes.
Michael,
Certainly if a company has less than a third of its spend under management that should be a priority. Sounds like a great opportunity for that company to augment its staff with experienced consultants who can attack unmanaged spend on a gain sharing basis -- no upfront cost or risk. I can recommend a great firm! :)
Doc -
I don't mean to be rude or flame, but there are so many flaws in this post that I do not know where to start. After the recommendation to use a should-cost model, your entire example and series of approaches devolves rapidly into hypotheticals that are (a) unrealistic as to to how an organization would be sourcing these commodities and (b) offer terrible suggestions on how to navigate market fluctuations -- i.e. oil moves 65%, don't worry about it, steel drops 60% but stand pat on the contract you suggested that locks me in for only a 20% decrease... to name a few issues
No need for more caveats, etc. and I don't care if you don't post this -- But you really should bounce these ideas off a real sourcing professional before posting something like this.
Surprised:
Better idea: learn to do math *and* read word math before calling someone an idiot, because, otherwise, you'll end up looking like the idiot you're calling me.
(1) I'm not flat out saying "if oil moves 65%, don't worry about it".
I'm saying: with respect to a particular (single) should cost model, only worry about commodities that make up a significant portion of your cost that you don't expect to move more than an amount that would have a material impact on the total cost.
Let's say oil is 4%. If oil moved, in your example, 65%, then the total impact on our hypothetical should cost model is 0.04 * 0.65 = 0.026 = 2.6%. Now, if
(a) this is enough of a fluctuation to worry about in the category under consideration and
(b) you expected that oil might move this much
then you would include oil in your "significant" cost components and index against it.
(2) I never said that you should sign a contract that locks you into a 20% decrease when steel drops 60%. Let's look at what I said.
I said if you agreed to a cost decrease (off of the total cost) of 1% for every 3% reduction in steel cost, you'd be okay. Let's do the math ... AGAIN. Steel makes up 34% of the cost. So, if steel decreases 3%, your total cost decreases 0.03 * 0.34 = 0.0102 = 1.02% Or, in terms that might be easier for you to understand, for every $1 decrease in steel, your total cost goes down $0.34. Or, for every $3 decrease in steel, your total cost goes down $1.02. I was advocating that for every reduction of 3% in steel, you require a 1% reduction in total cost. I will admit that my rounding off to a clean percentage point to keep the explanation simple would cost you 0.02%, or, if steel dropped 60%, it would cost you 1.2%, and that might be too much ... but that's an easy fix ... simply calculate out the percentages to a couple of decimal points.
(3) Are you telling me you're a *real* sourcing professional? That's scary! If you can't even follow the simple math I used in my example, I have to wonder what kind of wool sales people are pulling over your eyes when you are negotiating on behalf of your clients.
The method I was explaining is not flawed. But, as Barb pointed out, you have to know when it makes sense to use it and when it does not. I wouldn't use on strategic categories, and I wouldn't use it on high-spend categories where there are too many significant cost components that could fluctuate wildly, because it only works IF you account for all of the significant cost components, which include any minor cost components that could fluctuate wildly enough to be a significant cost percentage.
Also, you can't use it if you can't do the math. Because if you think a 1% drop in price in a raw material component should translate into a 1% drop in total cost, you're not going to get it right.
I wasn't calling you an idiot, I think you write great stuff. I just disagree that your example is practical.
Surprised:
You wouldn't be alone if you were (calling me an idiot). Seems I annoyed some consultants*1 who, no surprises here, make their living off of repeat category sourcing on categories which might be applicable to this method, which even Jason "The Prophet" Busch indicated can be quite appropriate when properly implemented (see his recent post on Spend Matters).
I never said that this particular example on a power transformer is "practical" since "practical" is open to interpretation. For an applied mathematician like me, it is "practical" since I can account for all major cost components AND market price fluctuations using a 7 variable formula tracked against 7 indices (one each for steel, labor, copper, oil, and transportation; one for exchange rate fluctuations; and one for average wholesale price to account for movements in margin). However, if you're a math-phobe, I can see how that might look quite little complicated, especially since you'd probably need, as Jason points out, integrated contract management, procurement, and payment systems that automatically polled the indices on the defined schedule, updated the pricing formula, compared the invoiced prices to the expected prices, accounting for all of the agreed upon variances, and then alerted you if the invoiced prices were off by more than some defined percentage of expected value (to account for rounding errors).
As a skilled enterprise software architect, this automation scenario is also "practical" for me. But if you're in an organization bereft of IT skills stuck with old, incompatible IT systems, it may not be "practical" until you update your systems.
This example was chosen to illustrate the potential power of the methodology and how it can be used on categories with more than one major cost component subject to (high) variability. If that's too much for you, or your systems, stick to categories with only one major cost component with high variability or categories whose market price fluctuations closely correlate with the market price fluctuations of a publicly traded commodity and define price escalations and de-escalations off of the stock price.
I'm bored today...aha! An argument!
Doc, it wasn't the clearest thing in the world that you were discussing a transformer contract. I somehow picked up that it was a raw material contract that you were willing to take a 1/3 share of a price decrease. Probably others did too.
Lots of big companies negotiate raw material prices even if they don't use much themselves and allow their suppliers to buy at the big company's price. That's what I thought you were writing about.
Now,one of my mantras is "commoditize your buys." The best solution here is to beat your designers into designing for industry standard power transformers rather than something custom. That makes transformer buying a commodity product with shorter contracts and more competition to control pricing. Then you don't even have to discuss supplier's costs.
Dick:
I hope this means you're stepping up to the plate and sending me a contribution for my upcoming Category Sourcing Guest Author series ...