The brokered patent market in 2015

We started writing these articles four years ago (see “Turning the spotlight on the brokered patent market”, issue 57, “The brokered patent market in 2013”, issue 63 and “The brokered patent market in 2014”, issue 69), because we believed that the entire patent market would benefit from more transparency. Real, data-driven information about pricing, technology areas, deals terms, the diligence process and other elements relating to the buying process can make buying patents less intimidating and enable companies to start participating in a market that is otherwise clouded by obscurity. We want to use the information that we have collected to help grow the market.

This transparency needs to be unbiased, delivering both good and bad news. Recently, some sobering developments for rights holders have received a great deal of press – most notably Alice Corp Pty Ltd v CLS Bank Int’l, but also five other Supreme Court cases with unanimous (nine-to-zero) holdings. The US courts are also shifting damages law towards apportionment-based thinking for standard-essential patents (see Microsoft v Motorola and Ericsson v D-Link), and rights holders in the United States are feeling the pressure. Similarly, stocks of companies that focus on IP monetisation (eg, Acacia, Pendrell and Inventergy) are generally down, and Intellectual Ventures (IV) is buying substantially less than it was before. Funding sources for monetisation in the United States are drying up too, with the notable exception of the Fortress Investment Group. However, while this may sound ominous for the future of the patent market, our data shows that the news is not all bad.

The 2015 brokered patent market described by this article (June 1 2014 to May 31 2015) is still quite robust. Packages offered for sale accounted for $1.1 billion in buying opportunities and we estimate that the actual amount transacted in sales amounted to $233 million. The number of packages listed remained basically flat, at 566; however, packages included more total US-issued patents and more total assets on offer. Also, packages appear to be selling at a higher rate compared to the 2014 market rate. In addition, brokers appear to be succeeding at selling packages for less than $250,000, an indication of greater market efficiency. We have seen buyers experimenting with some interesting new models for buying. For example, earlier this year Google held a reverse patent auction and purchased a number of patents for between $3,000 and $250,000 each. Even packages in technology areas that are affected by Alice are still selling above the market rate. While there has been a decline in business process packages, software packages are still selling above the market rate. Overall, there has been a drop in total dollar value of the market, but it has by no means fallen off a cliff.

Trends that we saw last year continue to be apparent. Non-practising entities (NPEs) and corporations are still buying; aggregators are picking up more deals; and software is still hot, even with the negative case law. Before turning to the details of the 2015 market, it may be helpful to get a sense of the patent market over the past few years.

Figure 1. Cumulative sum of asking prices ($ million) – brokered and private markets

How big is the market?

This year, we looked at the size of the patent market in a new way, examining both private and public deals and determining the total dollar value. Note that our visibility into private deals is limited to deals in which we have had work. That said, the dollar value of the market is surprisingly large and diverse.

Further, sales remain robust with a cumulative $1.7 billion of sales tracked, based on asking prices.

Our database tracks the status of 64,105 patent assets (37,099 US-issued patents), spanning $6.9 billion of asking prices and nearly $2 billion in sales. The assets are found in 2,029 packages (279 private and 1,750 brokered). We maintain a technology taxonomy with 18 general categories and 106 sub-categories, and each package is categorised into a single taxonomy. We also track asking prices, bidding dates and clients’ specific diligence decisions.

The rest of the article provides an overall sizing of the market, following the flow of a typical purchase process. It covers sourcing, asking prices, diligence steps, purchase closing and litigation, and concludes with our market-size estimate.

Patent brokers

As in previous years, our analysis focuses on the brokered patent market because it is open to all buyers. We work with many brokers and assist rights holders that are looking for the right broker to help sell their assets. Brokers bring various useful skills to the table, including:

• providing an initial filter for patent assets;
• selecting viable sellers and providing some certainty that the seller is willing to sell;
• screening patents and identifying important patents and claims;
• providing pricing guidance;
• providing guidance for sellers with regard to sales terms and timelines;
• defining a process for diligence, bidding and sales;
• developing more evidence of uses (EOU); and
• providing tougher negotiation on pricing.

Figure 2. Cumulative sum of asking prices of sales ($ million) – brokered and private markets

Brokers with five or more packages

The number of brokers remains about constant at 60 (last year it was 58). With two more brokers and fewer overall packages, brokers are listing slightly fewer packages on average. Some brokers continue to bring the majority of the packages to market: 15 brokers brought 10 or more packages to market, while 80% of the packages were brought by brokers which brought five or more packages to market. The top four brokers accounted for 34% of listed packages (last year 33%). As in previous years, we continue to see little technology specialisation among brokers, with the exception of some affiliated with semiconductor reverse-engineering houses and a few which focus more on hardware across multiple technology areas.

As shown in Figure 3, while a few brokers were particularly successful (circled in green) or unsuccessful (circled in red), those which brought more packages to market did not show a higher sales rate. Some brokers are clearly struggling. Unsurprisingly, brokers bringing the most deals to the market approached the industry sales rate. We used the 2014 calendar year for this analysis to allow sufficient time for sales to close and be recorded.

Figure 3. 2014 broker sales rate by number of listed packages (calendar year)

At 566 packages (556 last year), the brokered patent market has neither expanded nor contracted (see Table 2). We have benchmarked our deal flow with other large corporations and defensive aggregators, and it appears that we are receiving the same number of, or more, brokered packages, so we are confident that our numbers reflect the market. A significant change has been the number of US-issued assets per package, with an increase of 43% in these for almost the same number of packages. We believe that brokers continue to focus on US assets to increase the likelihood of a sale.

Technology distribution

Overall, the market continues to present packages from a broad and varied set of technologies. With increased diversity and a healthy number of packages, assets are available in almost any high-tech category. When we receive a package, we use the package materials to categorise it according to our taxonomy of technical areas. It is a two-tiered classification, with 18 general technical categories and 106 sub-categories. As seen in Figure 4, the distribution of general technologies is heavily skewed towards software and cloud computing. Surprisingly, decisions on software-related technologies in Alice and related rulings from the Patent Trial and Appeal Board (PTAB) did not result in a reduced number of software-related packages. Business process related patents, such as those in State Street and most likely affected by Alice and other subject-matter related court decisions, were still strongly represented, even though the case law clearly puts those patents in jeopardy. Alice’s impact on the sales rates is discussed below.

Figure 4. Package distribution by technology category

Figure 5. Frequency of package sizes (total assets)

Package sizes

Last year we saw a shift to smaller packages and the distribution also held true this year – except for a sharp decline in single patent packages (down nearly 30%) (Figure 5). This decline accounts for some of the increase in the average number of assets per package (15.63, up from 12.63). The remaining increase in assets per package is due to an increase in packages with between 25 and 199 assets.

Pricing

What is the price of a patent? What is a fair price? Am I getting a good deal? These are the questions that we continue to address with our pricing analysis. We use this analysis not only to help our clients buy and sell patents, but also to provide market pricing based valuation – if you know the asking market price of an average patent, you can build the model to price a specific patent. Importantly, this avoids having to determine an imputed royalty rate for the valuation.

Asking prices in 2015 continued to drop, with both per asset and per US-issued patent asking prices dropping $57,000 and $60,000 respectively. This represents a 20% drop on 2014. Due to a methodology change, a straight comparison to the table in last year’s paper shows an additional $23,000 drop. Table 3 provides a comparison view of last year’s numbers using our current methodology.

Figure 6. Average asking price per asset by month (top and bottom 5% removed)

Throughout 2015, we kept hearing about the market dropping off quickly, so we were concerned that when calculating the asking price per asset, we might overlook changes within the year. We investigated further and found that this was not the case (Figure 6). While there was a discernable drop in the monthly market size due to fluctuations in package size, there was no drop-off in asset asking prices. The average asking price per asset has remained fairly consistent throughout the market year.

Figure 7 shows the distribution of asking prices and a continued focus on packages priced in the $250,000 to $2 million range (into which 63% of all packages fall). Here, brokers make a healthy commission while keeping purchases within buyers’ budgets so that no one deal ties up the entire budget.

Figure 7. Distribution of package asking prices – top and bottom 5% removed

This year, we also began tracking the sub-$250,000 price range separately. This range is interesting in that it represents relatively low margins for a broker. At a 20% commission, a maximum $50,000 commission is possible; when one takes into account the overall low sales rate of packages (more on this below), it is clear that brokers are finding ways to lower their costs when bringing packages to the market. For example, there is rarely an EOU for packages priced below $250,000. In buying these types of package, we advise clients to scale down the resources they spend on diligence and negotiating the patent purchase agreement unless the buyer has significant plans for the patents.

Packages with pricing guidance

Eight-four percent of packages came with pricing guidance – almost the exact same proportion as last year (83%). Packages without pricing guidance continue to put themselves at a disadvantage, with sales rates for these lagging behind the market. While the percentage of packages with guidance remained constant, we saw a drop in the level of precision in the guidance provided; 40% gave exact numbers in the 2014 market, but only 26% gave exact numbers in the 2015 market. It could be that, with prices dropping, brokers are less certain of a price. However, we know that clear pricing guidance helps buyers to make decisions – without guidance, the risk of no decision is higher (meaning no sale), simply because the seller is signalling a lack of understanding of where the market is. Clear pricing sends signals that the seller has some understanding of the market and that a deal is thus more likely to close. Additionally, pricing targets help buyers to initiate appropriate levels of diligence.

Per-asset pricing by package size

Figure 8. Per-asset asking price by package size

We also reviewed how pricing changes based on the number of assets in the package (Figures 8 and 9). Unsurprisingly, on average, per-asset pricing drops considerably as the size of the package increases (from almost $400,000 to just over $50,000). Taking the time to find and highlight the key patents and group those into smaller deals is likely the best course of action. However, there is an alternative view when considering large portfolios – in some cases, there may be only a few key patents to highlight. Despite the premium for these key patents, breaking them out may leave a package with a lot of undifferentiated assets which may not sell. In this situation, you may get a better return by selling a greater portion of your assets at a lower rate. This calculation can be difficult and requires up-to-date knowledge of the market and in-depth knowledge of the specific portfolio.

Generally, per-asset pricing is relatively constant in the $250,000 to $10 million asking price range, with the low being $169,000 and the high $227,000 (Figure 9). Pricing per asset drops for packages below $250,000, indicating higher-risk or lower-value patents (eg, no infringement, recent priority dates or almost expired patents). At the other end, pricing in the $10 million-plus packages remains well above the market average and represents deals that are either premium assets or highly unlikely to close at those per-asset price points.

Figure 9. Per-asset asking price by package asking price

Figure 10. Distribution of package asking prices – with EOU

Figure 11. Distribution of package asking prices – without EOU

Broker asking price and impact of EOUs

This year, we see a pricing premium for packages with EOUs of approximately $60,000, up to a 34% premium (Table 4). Figures 10 and 11 show the pricing premium for EOUs with a clear shift in per-asset asking prices when EOUs are provided. The drop in asking price from 2014 to 2015 for packages with EOUs was only 4%, compared to the 20% drop across all packages. In the 2013 market, we saw a 27% mark-up for packages where EOUs were provided; in the 2014 market, this mark-up was less than 5%.

We saw a drop in the percentage of packages with EOUs (37% compared to 47% last year), which is surprising given the pricing premium. We continue to see brokers asking sellers for upfront fees to cover some of the costs of preparing the offering package, including EOUs.

Asking price by tech category

The average asking price per US-issued patent by tech category was calculated for every tech area with at least five packages on the market (Figure 12), although we removed the top and bottom 5% of total market data points. Tech categories continue to be strong drivers for asking price, as the category with the highest asking price per asset – social networks – demanded 322% of the asking price per asset compared to packages in the components category. Generally speaking, technology areas relating to internet computing still have the highest asking prices, despite potential Alice implications.

Technologies affected by Alice

The listing rates packages in technology areas affected by Alice and PTAB decisions do not seem to have changed since the ruling (Figures 13 and 14). Figure 13 compares the listing rates for both Alice affected and non-affected packages, which have followed the same general trend. While listing rates have not changed, sales rates are declining (discussed below), but not by as much as might have been predicted.

Figure 12. Average asking price per asset by technology category

Figure 13. Package listings rates per quarter by Alice impact (moving average)

Key diligence data

We would like to end the use of the phrase ‘low quality’ with regard to patents when discussing potential patent purchases. We often hear that there are junk, low-quality or weak patents on the market. Clearly there are some patents that we can objectively agree are poor. However, the majority of patents are never tested for quality metrics (eg, enforceability), as it is simply too expensive to do so; the objective quality metrics were never tested.

Figure 14. Percentage of total package listings affected by Alice

An important goal of any effective patent purchasing programme is to reduce the cost of finding the patents that fit your business needs. Eliminating patents because they do not meet business criteria (before any other diligence is performed) is the most cost-effective way to find patents that fit your needs. However, this means that the quality of the patents is rarely reviewed. After a package has been identified as potentially interesting to a client (eg, wearables), the next step is to see whether the technology described is something of real interest to the client. The question could be posed like this: if we assume that the patent is otherwise perfect, would we still buy it? Often (72%), the answer is no – the patent may simply not be relevant to the specific client (eg, the wearable may be applicable to watches, but not fitness trackers). Of the remaining 28% that made it through the technology filter (Table 5), 43% were passed over because the applicable market was too small (eg, insufficient damages, wrong group of potential infringers). The remaining 57% (16% of the total) underwent detailed tested of the assets.

Some areas where clients may have concerns rarely cause a package to be eliminated – for instance, if the bid due date is too soon, brokers can simply move it back. Pricing is also rarely a reason to pass because:

• brokers tend to know what a reasonable price for a package is, so brokered packages tend to be priced reasonably; and
• by the time clients decide to buy, they also have a good idea of a reasonable price to pay for the assets.

Sales

This year’s sales rates appear to be up slightly up on last year’s. Anecdotally, the market sounds as if it has fallen off a cliff, but the data says otherwise.

For sales data, we have to switch to a different data set, which includes 1,582 packages, 352 of which are identified as sales, and which is measured on the calendar year. This sample set includes packages that were analysed in our previous papers and goes back to packages listed as early as 2009. Our methodology uses the US Patent and Trademark Office (USPTO) assignments database to identify sales (if at least one patent in a package is found to have a sale assignment, that package is treated as sold). We improved our sales analysis by changing to an actual execution date, rather than a recorded assignment date. The data is limited to packages received by May 31 2015 and to sales recorded with the USPTO by August 20 2015.

Figure 15. Cumulative sales by years from package listing date

We continue to see an extended sales process and, in aggregate, the results remain similar to those from last year. However, sales rates and time to sell have stabilised and even improved somewhat.

The market continues to be robust, but tough. Our 2014 sales rate is currently at about 18.4%, well ahead of 2013’s rate of 14.9% (when viewed at a comparable time on market). The overall 2013 sales rate is down from the predicted result (27% versus 32%), reflecting a further drop-off in late sales (sales more than 18 months from the listing have dropped). The 51% sales rate from 2009 continues to be the high-water mark, and the market shows no indication of returning to that rate. Bear in mind that this sales data necessarily lags behind the actual market (by up to 18 months).

As we discussed in last year’s paper, the period after listing in which a package has a realistic chance of selling has continued to shorten. For packages listed in 2013, only 0.5% of the sales occurred in the third year of listing. This is a significant drop on 2012 (1.5%) and 2009 (12.3%). We expect 2014 to reflect a similar low sales rate at more than one year from listing. We believe that buyers have established their formal evaluation processes and are making decisions rapidly.

Our estimated sales rate for packages listed in calendar year 2014 was calculated as illustrated in Figure 15. So far, packages listed in 2014 are selling at a greater rate than those listed in 2013, and we therefore scaled up the actual 2013 first-year sales rate. The sales rates from packages listed in 2011-13 were used to estimate how quickly sales would fall off in subsequent years.

Sales by package size

We analysed the sales rate based on the size of the package listed and found that the highest sales rate occurred for packages in the six to 10 asset range (Table 6). Upon initial inspection, it appears that the largest packages actually do better, but we believe that this is due to the sales identification methodology. More sales are identified in larger packages because if any asset changes hands, the package is considered sold. We do not account for buyers cherry picking from large packages. That said, one would expect to see a much higher sales rate in the 51 to 100 asset range, using our methodology. We did see a higher rate in the 101 to 200 asset range (not shown), because buyers were cherry picking.

Sales by receipt date

We also analysed how quickly packages sell to estimate how much time buyers have to bid. While there is no pressure to bid quickly on undesirable packages, we wanted to focus on packages that did sell in order to calculate when buyers need to make their decision. Figure 16 shows that 80% of sales occurred within 11 months of the receipt date of the package. Some buyers are able to move very quickly, as is evident by about one-third of the packages selling in the first four months. Accelerated decision making is a buying advantage.

Figure 16. Cumulative percentage of sales by months from receipt date (2014 listings)

Sales by EOU provided

Anecdotally, we hear buyers saying that broker EOUs are not helpful, but the data says otherwise. As discussed in the pricing section above, a broker providing an EOU shows an increase in asking price of between 25% and 34%. However, this is not the only advantage that EOUs confer – they also increase the chance of a sale taking place. Packages with EOUs accounted for more than half (51%) of the sales of packages listed in 2014 or 2015. This is a distinct advantage, considering that the likelihood of an EOU being provided in this dataset was only 36%. Putting these numbers into context, packages with EOUs are 41% more likely to sell, while packages with no EOU are 23% less likely to sell than the average.

USPTO assignments

Across the sales we have tracked, regardless of year of sale, the average time between recorded and executed dates was 54 days. Based on our experience, we feel that anything over three months’ delay seems longer than necessary. Figure 17 shows the number of months from assignment to recording; 80% of sales are recorded within three months.

Figure 17. Number of months’ delay in recording the assignment after the executed date (all sales)

Impact of Alice on sales

Figure 18. Percentage difference between Alice-affected sales rate and total market sales rate

The Supreme Court handed down its ruling in Alice on June 19 2014 and a number of related PTAB, Federal Circuit and district court cases have since followed. We usually look at sales based on calendar years to give packages adequate time to sell before our analysis. However, we decided to take an early look at the 2015 market year because Alice coincides with the beginning of this period (Figure 18). While it is too early to know for sure, the trends look interesting.

We took our 106 tech sub-categories and labelled each as either Alice affected or non-Alice affected. Thirty-one sub-categories – including most software, business processes, social networking and advertising – were considered to be Alice affected. Then we compared the sales rates of the Alice-affected areas for the market years 2013, 2014 and 2015 (June 1 of the prior year to May 31 of the 2013 market year) to the respective total sales rates. It is no surprise that Alice-affected packages were selling at above the market rate before the decision; 15% better in the 2013 market and 20% better in the 2014 market. However, we were surprised to see that they are still selling quite well, at 10% better, in the 2015 market. While sales are not nearly as good as they were in 2014, the early data still shows above-average sales and demonstrates that Alice has not destroyed the market.

We also looked at the relative sales rates for business process and financial patents. This encompassed seven sub-categories, including business processes (specifically e-commerce, payments and traded instruments) and minimal hardware relating to payments (Figure 19). These sub-categories are down from selling at a rate 40% better than the 2014 market total sales rate to 45% below the 2015 market rate. As the business process packages are heavily Alice affected and the drop is so significant, this seems to account for much of the fall-off in sales. It seems as if Alice has limited much of its effect on the market to business software and has not had as big an impact on software technology as a whole.

Figure 19. Percentage difference between financial tech sales rate and total market sales rate

Sellers

We now turn to who is selling patents (Figure 20). We found that, as expected, the sales occurring from January 1 2014 to August 20 2015 (the last date on which assignments were checked for this data) were mostly by operating companies, 71% up from 62%. However, sales from individual inventors dropped significantly to 8%, down from 21%. There was also an increase in NPE sales to 16%, up from 11%. NPE buying activity also increased, so it appears that NPEs are buying more patents rather than selling off their assets. Potentially missing from these numbers are sales from defensive aggregators; we believe that many of these take place privately.

Figure 20. Distribution of seller type by sale year

Thirteen operating companies and two NPEs were part of the 15 entities that sold more than one package from January 1 2014 to August 20 2015. These sales accounted for 23% of sold packages, 41% of sold assets and 46% of sold US-issued patents. As we discussed in our article on IV (“What’s inside IV’s patent portfolio?”, issue 66), having cross-licences (or a licence on transfer) in place can substantially reduce a company’s exposure to patents from regular sellers. The repeat seller list should be a focus for any cross-licensing strategy (Table 7).

Buyers

For our analysis of buyers, we looked at the same timeframe as we did for sellers: January 1 2014 to August 20 2015. NPE purchases rose to 42%, up from 38%; as did defensive aggregator purchases, which increased to 21% up from 13% (Figure 21). Operating companies dropped a corresponding 12%, purchasing 34%, down from 46%. As we predicted in last year’s report, the increase in NPE activity is likely due to IV’s resumption of purchases for its Invention Investment Fund 3.

During this period, 108 buyers purchased 206 packages and 15 buyers purchased multiple packages. Repeat buyers purchased 54% of the packages sold, with 36% purchased by the top three buyers: IV, RPX Corporation and Allied Security Trust (Table 8). In fact, IV purchased 19% of packages, 25% of assets and 25% of US-issued patents on its own.

Figure 21. Distribution of buyer type by sale year

Litigation and inter partes reviews

We analysed all of the packages in our database with asset lists to discover how the assets were being used after we received the package from the broker. Only 2.1% of all packages contained an asset that was litigated after we received the package from the broker – regardless of whether the package was sold. Sold packages were 65% more likely to contain litigated assets, at 3.5%. Unsold packages were 20% less likely to have litigation, at 1.7% after receipt of the package.

We also looked at the rate of inter partes reviews. Of all packages tracked, we saw only seven with inter partes reviews filed after we received the package, a 0.47% inter partes review rate – regardless of whether the package was sold. Sold packages were 22% more likely to have an inter partes review, a 0.58% inter partes review rate. Unsold packages were 7% less likely to have an inter partes review, a 0.44% inter partes review rate. Across these seven packages, 12 unique patents were subject to inter partes reviews 14 times; two were filed in 2014 and 12 were filed in 2015. This indicates that, as expected, inter partes reviews are on the rise and brokered patents are being used in assertions. Because inter partes reviews are still relatively new, we expect this trend to be greater than the data currently indicates.

Full market size

The market has shrunk again in 2015, but not as much as one might have feared given the news. We estimate that actual sales from June 1 2014 to May 31 2015 were $233 million (down from $260 million last year).

Compared to previous years, we have adopted a number of methodology changes using:

• executed dates in our sale identification process;
• new insights into the time it takes for a company to record a sale; and
• an improved average pricing methodology.

These in turn have enabled us to simplify our total market estimation method. We believe that our new method of using the actual observed sales that were executed in the 2015 market year timeframe and their asking prices is both simpler and more accurate. If a sale was observed with no pricing guidance, the average asking price per asset of $189,880 was multiplied by the number of assets to determine the expected asking price.

In the 2015 market year, 136 sales were identified, accounting for $342 million in total asking price. We know that some of the sales in that period have not yet been recorded, estimated at approximately 3%, so we reduced our standard 35% discount between asking price and expected selling price to 32%. Thus our expected total market size for the 2015 market is $233 million. In our analysis last year, we estimated the 2014 market at $260 million, so the market has declined by approximately 10%.

Using an average commission rate of 20%, the revenue from this market for brokers is $47 million per year. We back-tested the market size by estimating the average loaded labour rate per broker ($300,000 a year), resulting in 157 full-time equivalent brokers. Assuming that three brokers work in each brokerage, this results in approximately 52 brokerages. Our data shows 61 brokerages, which suggests that there are even fewer brokers per brokerage, or that brokers are doing other things, such as consulting. Additionally, each brokerage brings about 11 packages to the market per year. However, our data shows that a few brokers bring many packages to the market, with the majority bringing a few packages.

Opportunities, conclusions and reflections

We started writing these reports nearly five years ago to help inform the market and we have succeeded far beyond our most optimistic predictions. Opportunities for operating companies needing to fill a niche remain abundant, with nearly 50 packages entering the market every month. More efficient pricing offers companies the opportunity to pick up patents in the sub-$250,000 range.

Some best practices for sellers across the years include the following:

• Generally segment your assets for sale into smaller groupings to help sales;
• Be willing to provide clear pricing guidance; and
• Where possible, develop and share EOU materials with the buying community.

For operating companies, the market data continues to support the need to plan for handling future sales of patents by operating companies to NPEs. Cross-licensing and licensing on transfer are just some of the powerful approaches for reducing future NPE risks.

We look forward to continuing to help our clients transact millions of dollars of patents in 2016 and will continue working to improve our data collection and analysis practices to answer additional questions in next year’s report.