Structuring Royalty Payments to Mutual Advantage

Crispin Marsh, Managing Director, SCP Technology and Growth Pty Ltd

• Introduction
• The "25% Rule"
• Royalty Rate Structures
• Constant Royalty
• Positively Correlated Royalties
• Negatively Correlated Royalties
• Conclusion
• Selected Bibliography

Example Microsoft Excel Spreadsheet

This publication is provided with an example spreadsheet file. Please download the file, and try changing the Sales and Royalty amounts to view the effect on sales, royalties and profits.

• Download the Excel file
• Download the file in Zip format

The literature in the licensing field typically suggests that the most common form of remuneration for the grant of a licence will include a royalty, i.e. a payment made by the licensee to the licensor which is proportional to the amount to which the licensee has used the licensed technology or has produced the licensed product. This royalty is frequently defined as a proportion of the net selling price of a product produced by the licensee using the licensor's technology. In other licenses the royalty may be calculated as a fixed sum per licensed product sold.

Frequently a licence agreement will also provide for the royalty payment to be supplemented by other payments such as front-end fees, consultancy fees, minimum royalties and component supply fees.

There has been a considerable amount written about how the parties might settle on a "fair" royalty in any given situation. The complexity of the subject can be seen from the article by Tom Arnold entitled 100 Factors in Determining a Royalty. This theoretical literature has been supplemented by surveys conducted by investigators associated with LES and published in the LES quarterly journal Les Nouvelles.

In this paper another issue relating to royalty rates will be explored. This is the structuring of the royalty rate so that it may vary through the life of the licence in accordance with production quantity, time or some other factor. It will be shown that in many cases a "fair royalty" provision in a licence is one that provides that the royalty rate is not fixed throughout the term of the licence, but rather may vary in accordance with an agreed formula over that term. Doing this can reduce the risks that a royalty, which seemed fair at the time the licence was entered into, will, in practice be unfair to one party or the other.

Before exploring this issue further it is helpful to review one particular approach to determining a "fair" royalty as this method will be used to illustrate the advantages to be gained from varying the royalty over the life of the licence.

Among the techniques for determining a "fair" royalty is the so-called "25% rule". This "rule" suggests that a licensor should receive 25% of the extra profit derived from the licensee's use of the licensed technology. There are a number of points that should be borne in mind when considering the "25% rule". First and foremost it must be remembered that it is but one of a number of alternative approaches that may be used to arrive at an agreed quantum of consideration for the grant of a licence. To this extent it is a misnomer to call this technique a rule. It is nothing more than an aid to reaching agreement between a licensor and a licensee; despite these reservations the author, and it is believed many other licensing executives, have found the technique helpful, and it often accords with industry practice.

Secondly the "rule" needs to be applied over the estimated life of the licence. One needs therefore to think in terms of the Net Present Value (NPV) of the estimated royalty stream, taken together with other consideration to be received by the licensee, over the estimated life of the licence as being equal to 25% of the NPV of the estimated "extra" profit that the licensee will derive from use of the licensed technology. This allows for the fact that a licensee may well make a loss in the first few years of the licence despite which the licensor will be receiving royalties. A corollary is that in the later years of the licence the licensee will recoup those early year losses through receiving more than 75% of the annual "extra" profit derived from use of the licensed technology. Obviously, in order to carry out the NPV calculation, it is necessary to develop a business plan to model the licence over it's life. It is acknowledged that there are inherent difficulties and inaccuracies in doing this. It is suggested that the advantages of developing a coherent business plan and basing one's negotiations on this plan is preferable to merely accepting an "industry standard" or, worse still, arbitrarily picking a number!

Thirdly it must be made clear that it is not suggested that the consideration clause should provide directly that the royalty is 25% of the licensee's profit. There are many reasons why this would be so, foremost among these being that "profit" is an ill defined concept open to manipulation. Rather, a royalty rate is selected, normally as a percentage of the net selling price of the licensed product, such that over the life of the licence the licensor receives an appropriate proportion of the "extra" profit made by the licensee.

The term "extra" profit is used as the licensee may already be selling a product which the licensed technology improves. The "25 % rule" calculation only applies to that proportion of profit that is made as a result of the adoption of the licensor's technology. A simple example illustrates the concept. Imagine that an inventor finds that throwing a handful of gum leaves into a blast furnace from time to time improves the process to such an extent that the steel produced from the blast furnace is cheaper by $2.00 per tonne. It can be assumed that the steel producer will sell the steel at it's prevailing price and therefore a saving of $2.00 per tonne will provide the steel manufacturer with an additional $2.00 per tonne profit increment over and above its normal profit margin. The inventor's royalty, applying the "25% rule", would thus be $0.50 per tonne, i.e. 25% of the incremental increase in profit, not 25% of the steel manufacturer's total profit.

Typically simple licence agreements provide for a single royalty during the life of the licence. This might be, say, 5% of the net selling price or, alternatively, say, $0.50 per product. In slightly more complex licenses these alternatives may be combined by saying that the royalty shall be 5% but in no case shall it be less than (or more than) $0.50 per product.

Varying royalty rates are generally introduced in their simplest form where the royalty changes in different circumstances. The royalty may, for instance, be halved in those countries where there is no patent protection but where licensed know-how is used. Alternatively the royalty may be different in two or more specific fields due to the varying profit margins that apply in those respective fields. Such a situation might apply where a diagnostic test is used in the food industry and in the health care industry. The profit margins in the food industry are typically much smaller than those in the health care industry and royalty rates in those industries typically reflect this.

There is another more complex issue to be considered; that is whether the royalty rate on a licensed product in a particular industry and country should remain constant throughout the life of the license, or alternatively, whether provision should be made for it to vary. The variation may be upwards as some other factor increases, i.e. there is a positive correlation with the factor, or it may be varied downwardly as that factor increases, i.e. there is a negative correlation with the factor. If, for instance, the factor is time, then a positive correlation between time and royalty would mean that the royalty would increase progressively through the life of the licence. A negative correlation would mean that the royalty would decrease over the life of the licence.

One might ask why there would be value in drafting a licence providing for either a positive or a negative correlation between royalty rate and some other factor. The short answer is that there are often circumstances in which such a provision allows the licence to have a flexibility which makes it more likely that the royalty will be "fair" over the life of the licence irrespective of whether or not the original business plan proves to be correct. The factors that might be relevant in selecting either a constant royalty or one with a correlation to another factor are considered further in this paper. The "25% rule" can be used as a means of illustrating whether the result of the use of different royalty structures is "fair" or not.

If the technology being licensed is capable of being used commercially without any great expenditure of capital and if the future of the technology can be predicted with reasonable certainty then a constant royalty is an appropriate mechanism for compensating the licensor. A constant royalty also has the advantage of simplicity for accounting and auditing purposes.

Table 1

Table 1 shows a hypothetical business plan involving the licensing of a product which will yield a profit from the first year and which is expected to increase in sales over a period and then to stabilise until the end of the licence period. It can be seen that the business plan provides for a 5% royalty and that it is anticipated that the licensee's margin will remain substantially constant over the life of the licence. The discount rate used in the business plan is modest as the perceived risk in the use of the technology is relatively low.

It can be seen that the NPV of the royalty represents 25% of the NPV of the "extra" profit derived by the licensee (which in this case is in fact the whole profit derived from the sale of the product as it is assumed that the product is a new one for the licensee).

The outcome of this business plan is also shown in Charts 1 and 2 which respectively depict the sharing of the NPV of the "extra" profit between the licensor and the licensee, and the sales, royalty and licensee profit provided for in the business plan over the life of the licence.

Chart 1	Chart 2

While the business plan seems to provide for a "fair" royalty what happens if the outcomes from the commercialisation of the technology are not as predicted in the business plan? If the sales are actually double those postulated in the business plan but with essentially the same margin of licensee profit, then the new sales figures can be inserted into the spreadsheet as is shown in Table 2. It can be seen that in these circumstances the sharing of the extra profit is unchanged.

Table 2

The same outcome is observed if the sales are only half those expected from the business plan.

A situation such as that postulated in this example may arise where a fully developed technology is licensed to an established manufacturer in a market where the margin of "extra" profit from the use of the technology is substantially independent of sales volume. A manufacturing process improvement might be such a case, though in that circumstance the "extra" profit would only be the profit derived from the use of the process improvement and not the total profit from the resultant goods. This situation would also apply where the expected increase in margin due to economies of scale as production increases is expected to be offset by competition in the market place with the effect that the overall margin remains substantially constant.

There are a number of other scenarios that can be envisaged for this hypothetical business plan. What would happen if the licensee had to put in substantial capital in the early years of the licence? What would happen if it was expected that market maturity in the later years of the licence would erode the overall margin of profit for the licensee? What would be the effect of an up-front payment to the licensor? These and other likely occurrences, when taken together with the inevitable uncertainties in any business plan, result in more complex, and varying, royalty provisions offering a greater likelihood that a "fair" division of the "extra" profit from the new technology will result. These possibilities will now be considered in more detail.

Let us now consider the situation where a relatively undeveloped technology with a potentially very large market is licensed. In the business plan there is provision for the payment of a front end fee of $1M to the licensor and that the licensee will make substantial losses in the initial years of the licence agreement due to the conduct of R&D, manufacturing development, regulatory approval and the like. Due to the much greater risks inherent in the project the discount rate used in the NPV calculations is 30%. It is to be expected that the licensee's margin will be dependent upon volume and this is varied with sales. Table 3 shows the expected outcome at a constant royalty rate.

Table 3

The proposed royalty satisfies the "25% rule" for the sales volume contemplated by the business plan, however, it does not address the problem of what will happen if the sales volume is greater or less than that envisaged in that plan.

Table 4 illustrates what happens if the sales are only half those envisaged in the business plan. It can be seen that as sales decrease so the share of the "extra" profit is slanted in favour of the licensor being 47% of the "extra" profit.

Table 4

Conversely as the sales increase the share is weighted in favour of the licensee, a doubling of sales in this example results in the licensee receiving only 14% of the "extra" profit.

It can thus be seen that the selection of a constant royalty rate presents both the licensor and the licensee with inherently greater risks than are desirable. For the licensor there is the risk that he will fund the project and then find that the licensee receives the major proportion of the profit from the activity if the sales are low. While this may seem like a potential win for the licensee, in practice this is unlikely. Either the potential licensee will not enter into the licence at all or, if it has already done so when he realises that the market may be smaller that anticipated, it will not actively pursue the technology. On the other hand if the technology earns more than was anticipated then the licensee will have been disadvantaged.

The problem of the slanting of the sharing of the "extra" profit either to the licensee or to the licensor with changing sales volumes can be addressed by providing that the royalty rate will vary in accordance with sales. If the sales are low then the royalty will be low; if the sales are substantial then the royalty rate will be substantial. Let us now look at what happens if we have a royalty rate of 0.22% for sales up to $55 Million, 7% for sales between $55M and $110M and 17% for sales above $110M. At the sales figures postulated in the original business plan the NPV of the royalty is 25% of the "extra" profits shown in Table 5.

Table 5

If the sales are either halved or doubled then the "25% rule" is still adhered to as is seen respectively in Tables 6 and 7.

Halved Sales

Table 6

Doubled Sales

Table 7

The model that has been created is quite robust for other variations of sales volume and provides for a sharing of the risk that the sales will be higher, or lower, than expected. It may seem an extreme circumstance where the royalty varies so enormously with variations in sales volume. To some extent this is a reflection of the example that has been chosen in which there is a substantial up-front payment in a situation that demands the licensee invest substantial monies in the first few years of the project with the hope of earning large profits in later years of the project. In practice the author has been involved in the negotiation of licences where agreement was reached on royalty rates which varied substantially with sales. In one case the royalty rates were 4%, 7% and 12%. In another case there were 5 steps in the cascade and these involved royalties of 1%, 1.5%, 2%, 2.5% and 3%.

In this example the higher royalty only applies to that amount of the royalty which is above each of the threshold values at which the higher royalty cuts in. One could imagine a situation where the higher royalty rate applies to all sales once the threshold has been reached. The problem with using such an arrangement is that once the licensee jumps over a threshold his total royalty bill will jump substantially. There is thus a disincentive to move just above the threshold value. In the author's experience the effect of such a provision is generally disadvantageous for the licensor due to the disincentive it provides to the licensee.

Positive correlation of royalties is appropriate where there is a factor which will cause the profitability of working the licence to increase rapidly with an increase in some external factor. In the above example the high up-front payments postulated caused the profit to rise sharply with volume because that increasing volume allowed the amortisation of the up-front costs over more product. A similar positive correlation would occur in any situation where the profit margin increased with sales volume. One situation where this will often happen is where there is a substantial fixed cost that has to be paid in order to produce the licensed product. This fixed cost could be plant cost, advertising cost, or the cost of establishing a sales force. The same situation could also occur where the licensed producer had to contend in the early years of the new product with an entrenched competitor using an outdated technology who was prepared to sell at a loss in the hope of driving the new producer from the market. The licensed producer's margin will rise once the old manufacturer succumbs to the new technology and is driven from the market place.

There will be many other situations where such a positive correlation can be established between royalty and some other factor in the market place. If it were foreseen, for example, that competition from a party who might design around the licensor's intellectual property was the greatest threat then one could provide that the royalty rate was to be positively correlated with the selling price of the product. One could provide, for instance, that for an arms length selling price of less than $10 the royalty is 3%; between $10 and $15 the royalty is 4%; and above $15 the royalty is 5%. While this arrangement may be susceptible to a "transfer pricing" problem with an unscrupulous licensee, it does produce a "fairer" division of the "extra" profit between licensor and licensee.

In the case of the licence between Ron Hickman and Black & Decker in respect of the "Workmate" work bench it is understood that the royalty rate was linked to the scope of the patent claims granted. This is also a form of positive correlation; bigger claim scope provided bigger royalties.

The issue of positive correlation has to this point been discussed from a financial point of view. It is suggested that the positive correlation model can be very effective as a tool in one's negotiating toolkit to break impasses during royalty negotiations. It does two things simultaneously; it allows each party to see their suggested royalty applying for part of the time, and it avoids the possibility that one party will do very well from the technology while another does very poorly. One often sees, as an undercurrent in licence negotiations, that the outcome that is most unpalatable to an industry based licensing executive is that she licenses a technology from a university and in time sees her company paying substantial royalties but earning little. Conversely the university licensing adviser sits at the table with gritted teeth determined not to be accused, in a few years time, of "giving away" the university's technology because the company is earning heaps and paying the university a pittance!

The example described with reference to Table 7 deals with the situation where the sales may vary in quantity. It does not, however, deal very well with delays in the commencement of sales. If the commencement of sales is delayed then this will tend to skew the sharing of the "extra" profit in favour of the licensor. If it was felt during the licensing negotiations that the greatest risk factor in the business plan was that there would be a delay in entering the market, for instance due to delays in obtaining regulatory approval, then the royalty rate structure would need to be adjusted for this. One way to do this would be to provide that the royalty rate would fall depending upon which year of the life of the agreement the product was first commercialised. This is, of course, a negative correlation between years and royalty rate. Let us now consider when negative royalty rate correlations might be appropriate.

In the preceding section we looked at some of the factors which made it appropriate to raise the royalty rate in accordance with an increase in some other factor. There are a number of circumstances where a negative correlation between the royalty rate and another factor is appropriate to provide a "fair" division between a licensor and a licensee.

Imagine that a new product comes onto the market with a certain margin of profit and that as the market matures the licensee's profit margin is eroded. Such a situation can be exemplified with reference to Table 8.

Table 8

In this example a developed product is expected to be profitable from year one without substantial initial costs of a capital nature. The margin is however postulated to fall as the market matures. The "early adopters" are perceived as being prepared to pay a high margin (20%) for the product whereas the general market will not accept the product at the price that this requires and the maximum market penetration will only be accepted at a margin of 10%. In this case it may be in everyone's interest to have a royalty rate that falls as sales rise. In this way both the licensor and the licensee are rewarded by the price that the early adopter will pay but they also share "fairly" the lower price, and thus margin, that the bulk of the market will pay. This is achieved by providing a royalty rate which falls with sales volume, from 5.5% for sales up to $17 Million, to 3.5% for sales in the band between $17 and 35 Millions, and to 2.5% for sales above $35 Million. In this case the sales could be half of those predicted or double those predicted and in each case running the spreadsheet shown in Table 8 shows that 25% of the "extra" profit would still go to the licensee. By contrast a flat 4.4% royalty would produce the desired "fair" division at the projected sales volume but the licensee would receive only 21% of the "extra" profit if sales were halved and a bonus 29% if sales doubled. Thus the provision of the negatively correlated royalty rate has reduced the risk to each of the parties that the sales will vary from those predicted in the business plan.

It is interesting that one much more often sees in the literature reference to a negative correlation model, in the form of falling royalty rates with increasing sales, than a positive correlation model. This may be due to an expectation of falling margins or due to an expectation that in time the licensee's own innovations will be driving sales and keeping competition at bay. This arrangement also may appeal to licensors as being a positive incentive to the licensee to increase their sales and thus increase the total royalty even if it means that the licensee gets a greater than "fair" proportion of the profit from the higher sales level achieved.

In extreme cases the negative correlation provides for total cessation of royalties when total royalties paid equal a cap of royalty payments. It is hard to conceive of the logic driving such a deal and one is forced to question the relative bargaining strengths of the parties to such arrangements. It may be comforting to the licensee to feel that the royalty cannot exceed by more than a known sum the amount that they may have paid to purchase the technology outright in the first instance. In this case there may be little reward for the licensor taking the risk of hanging out for a royalty rather than selling outright at an early stage. The licensor appears in this case to be accepting the downside risk that there will be little or no royalty, but does not enjoy the upside risk that sales will be better than expected.

Other circumstances where a negative correlation may be desirable include the situation where the profit margin of a technology is tied to the price of a raw material. This would be the case where the market demand for the manufactured product is relatively elastic meaning that the increased cost of the raw material cannot be passed on to the customers for the manufactured product. As the raw material price goes down so the profit margin of the technology user goes up and therefore the royalty should also go up. Conversely if the raw material price goes up then the margin will be less and the royalty comes down. Another example occurs if a major issue in the licence negotiations is the time it will take to commercialise the technology. In this case there could be a negative correlation between the royalty rate and the year of the licence in which the royalty is earned. Thus one could provide for a royalty of 6% in years 1 and 2, and in each year thereafter the royalty reduces by, say, 0.3%. Another variation on this theme is to provide that the royalty rate is fixed during the life of the licence but that the longer it takes to get into production the lower that rate will be, e.g. the royalty will be 6% if commercial production commences in either of years 1 or 2 but will reduce by, say, 0.3% for each additional year required to get into commercial production.

If the circumstances warranted it there could be a royalty rate determination that had a combination of correlations which could be both positive, both negative or one negative and one positive correlation. The latter arrangement could, for instance, be a positive correlation with sales and a negative correlation with date of first commercial sale. This could be achieved by way of a provision that said there were three bands of royalty rates for different sales volumes but that all of these rates would be reduced for each year of delay in commercialisation.

It has been suggested that there is value in negotiating royalty rates to seek an outcome that reduces the major risks that the parties perceive in the postulated business plan. This may be achieved through a mechanism of correlating the royalty rate, either positively or negatively with another factor. This procedure may be helpful in breaking a negotiation impasse, or in overcoming a perception by one party that if the plan does not work out as expected one party will be substantially advantaged at the other's expense.

It is stressed that this technique is not applicable in every case and is no more than one implement in the licensing executive's toolkit.

• Licensing Technology, Drafting and Negotiating Agreements, Noel Byrne, 1994 Macmillan press. See particularly chapter 3 "Royalty Determination".
• 100 Factors Involved in Pricing the Technology Licence, Arnold White & Durkee, 1988 Licensing Law Handbook, Clark Boardman
• Eckstrom's Licensing in Foreign Operations, Updated by David M. Epstein, 1998, West Group. See particularly section 3.04 "Payments for the Licence"
• A Survey of Licensed Royalties, Stephen A. Degnan and Corwin Horton, Les Nouvelles June 1997.
• Licensing Practices, Business Strategy, and Factors Affecting Royalty Rates: Results of Survey, Daniel M. McGavock, David A. Haas and Michael P. Patin, Licensing Law and Business Report March-April 1991, Clark Boardman.
• Royalty Rates Conform to 'Industry Norms', Patrick H. Sullivan Les Nouvelles September 1994.
• New Companion to Licensing Negotiations, Licensing Law Handbook, 1996-97 Edition, Robert Goldscheider, Clark Bordman Callaghan.
• Patent Infringement Damages Awards 1982-1994, Julie L. Davis, Licensing Law and Business Report May-June 1995, Clark Boardman Callaghan.
• Georgia-Pacific Corporation v United States Plywood Corporation (1970) 166 USPQ 235 (United States District Court, S.D. New York)