The Coming Breakthroughs in the Global Supercomputing Race

In our August 31, 2015 article (“The U.S. Needs to Rejuvenate the Global Supercomputing Race“), we expressed our concerns regarding the state of the global supercomputing industry; specifically, from a U.S. perspective, the sustainability of Moore’s Law, as well as increasing competition from the Chinese supercomputing industry. Below is a summary of the concerns that we expressed:

  • Technological innovation, along with increasing access to cheap, abundant energy, is the lifeblood of a growing, modern economy. As chronicled by Professor Robert Gordon in “The Rise and Fall of American Growth,” U.S. productivity growth (see Figure 1 below; sources: Professor Gordon & American Enterprise Institute)–with the exception of a brief spurt from 1997-2004–peaked during the period from the late 1920s to the early 1950s; by 1970, much of today’s everyday household conveniences, along with the most important innovations in transportation & medicine, have already been invented and diffused across the U.S. Since 1970, almost all of the U.S. productivity growth could be attributed to the adoption and advances in the PC, investments in our fiber optic and wireless networks, along with the accompanying growth of the U.S. software industry (other impactful technologies since the 1970s include: the advent of hydraulic fracturing in oil & gas shale, ultra deepwater drilling in the Gulf of Mexico, as well as the commercialization of alternative energy and more efficient battery storage systems, as we first discussed in our July 27, 2014 article “How Fracking Saved the U.S. Economy“). This means that a stagnation in the U.S. computing or communications industries would result in an invariable slowdown in U.S/global productivity growth;

americanproductivitygrowth

  • The progress of the U.S. supercomputing industry, as measured by the traditional FLOPS (floating-point operations per second) benchmark, had experienced a relative stagnation when we last wrote about the topic in August 2015. E.g. in 2011, both Intel and SGI seriously discussed the commercialization of an “exascale” supercomputer (i.e. a system capable of performing 1 x 10^18 calculations per second) by the 2019-2020 time frame. As of today, the U.S. supercomputing community has pushed back its target time frame of building an exascale supercomputer to 2023;
  • At the country-specific level, the U.S. share of global supercomputing systems has been declining. As recent as 2012, the U.S. housed 55% of the world’s top 500 supercomputing systems; Japan was second, with 12% of the world’s supercomputing systems, with China (8%) in third place. By the summer of 2015, the U.S. share of the world’s top 500 supercomputing systems has shrunk to 46%, although both Japan and China remained a distant second at 8%. Today, the Chinese supercomputing industry has led an unprecedented surge to claim parity with the U.S, as shown in Figure 2 below.

Figure 2: China – Reaching Parity with the U.S. in the # of Top 500 Supercomputerstop500

Since the invention of the transistor in the late 1940s and the advent of the supercomputing industry in the 1960s, the U.S. has always been the leader in the supercomputing industry in terms of innovation, sheer computing power, and building the customized software needed to take advantage of said supercomputing power (e.g. software designed for precision weather forecasting, gene sequencing, airplane and automobile design, protein folding, and now, artificial intelligence, etc.). With U.S. economic growth increasingly dependent on innovations in the U.S. computing industry and communications network–and with China now threatening to surpass the U.S. in terms of supercomputing power (caveat: China’s HPC software industry is probably still a decade behind)–it is imperative for both U.S. policymakers and corporations to encourage and provide more resources for the U.S. to stay ahead of the supercomputing race.

Unlike the tone of our August 31, 2015 article, however, we have grown more hopeful, primarily because of the following developments:

  • Moore’s Law is still alive and well: At CES 2017 in Las Vegas, Intel declared that Moore’s Law remains relevant, with a second-half target release date for its 10 nano-meter microprocessor chips.  At a subsequent nationally-televised meeting with President Trump earlier this month, Intel CEO Brian Krzanich announced the construction of its $7 billion Fab 42 in Arizona, a pilot plant for its new 7 nano-meter chips. Commercial production of the 7nm chips is schedule to occur in the 2020-2022 time frame, with most analysts expecting the new plant to incorporate more exotic technologies, such as gallium-nitride as a semiconductor material. The next iteration is 5nm chips; beyond 5 nano-meters, however, a more fundamental solution to extend Moore’s Law will need to occur, e.g. commercializing a graphene-based transistor;
  • GPU integration into supercomputing systems: The modern-day era of the GPU (graphics process unit) began in May 1995, when Nvidia commercialized its first graphics chip, the NV1, the first commercially-available GPU capable of 3D rendering and video acceleration. Unlike a CPU, the GPU is embedded with multiple threads of processing power, allowing it to perform many times more simultaneous calculations relative to a CPU. Historically, the supercomputing industry had been unable to take advantage of the sheer processing power of the GPU, given the lack of suitable programming languages specifically designed for GPUs. When the 1.75 petaflop Jaguar supercomputer was unveiled by Oak Ridge National Laboratory in 2009, it was notable as Jaguar was one of the first supercomputers to be outfitted with Nvidia GPUs. Its direct successor, the 17.59 petaflop Titan, was unveiled in 2012 with over 18,000 GPUs. At the time, this was a concern for two reasons: 1) hosting over 18,000 GPUs within a single system was unprecedented and would doom the project to endless failures and outages, and 2) there were too few programming codes to take advantage of the sheer processing power of the 18,000 GPUs. These concerns have proven to be unfounded; today, GPUs are turning home PCs into supercomputing systems while Google just rolled out a GPU cloud service focused on serving AI customers;
  • AI, machine-learning software commercialization: Perhaps one of the most surprising developments in recent years has been the advent of AI, machine-learning software, yielding results that were unthinkable just five years ago. These include: 1) Google DeepMind’s AlphaGo, which defeated three-time European Go champion Fan Hui by 5-0 in 2015, and finally, the world Go champion Ke Jie earlieir this year, 2) Carnegie Mellon’s Libratus, which defeated four of the world’s top poker players over 20 days of playing, and 3) the inevitable commercialization of Level 5 autonomous vehicles on the streets of the U.S., likely by the 2021-2025 time frame. Most recently, Microsoft and the University of Cambridge teamed up to develop a machine learning system capable of writing its own code. The advent of AI in the early 21st century is likely to be a seminal event in the history of supercomputing;
  • Ongoing research into quantum computing: The development of a viable, commercial quantum computer is gaining traction and is probably 10-20 years away from realization. A quantum computer is necessary for the processing of tasks that are regarded as computationally intractable on a classical computer. These include: 1) drug discovery and the ability to customize medical treatments based on the simulation of proteins and how they interact with certain drug combinations, 2) invention of new materials through simulations at the atomic level. This will allow us to build better conductors and denser battery systems, thus transforming the U.S. energy infrastructure almost overnight, and 3) the ability to run simulations of complex societal and economic systems. This will allow us to more efficiently forecast economic growth and design better public policies and urban planning tools.

Why China Will Not Cut Rates Any Further This Year

In response to a slowing property market, lower consumer spending growth, and a slowing global economy, the People’s Bank of China (PBOC) has cut its one-year policy rate five times and its reserve requirement ratio three times over the last 12 months. Last November, the PBOC’s one-year policy rate sat at 6.00%–today, it is at 4.60%. Moreover, the PBOC’s cut in its reserve requirement ratio–from 20.0% to 18.0% since February–has released more than $400 billion in additional liquidity/lending capacity for the Chinese financial system.

I believe Chinese policymakers will maintain an easing bias over the next 6-12 months given the following:

  1. As I discussed a couple of years ago, a confluence of factors–including China’s debt build-up since the 2008-09 global financial crisis, slowing population growth, as well as natural limits to an export- and CAPEX-driven growth model–means China’s real GDP growth will slow to the 5%-8% range over the next several years. Consensus suggests that China’s real GDP growth will be lower than the official target of 7% this year. Given China’s significant debt build-up since the 2008-09 global financial crisis, policymakers will need to do more to lower lending costs and to encourage further lending as global economic growth continues to slow;
  2. Most of the debt build-up in China’s economy over the last 7 years has occurred within the country’s corporate sector–with real estate developers incurring much of the leverage. In other words, both real estate prices and investments are the most systemically important components of the Chinese economy. While real estate prices and sales in Tier 1 cities have been strong this year, those of Tier 2 and Tier 3 cities have not yet stabilized. This means policymakers will maintain an easing bias unless Chinese real estate sales and prices recover on a broader basis;
  3. Chinese credit growth in August met expectations, but demand for new loans did not. Real borrowing rates for the Chinese manufacturing sector is actually rising due to overcapacity issues and deteriorating balance sheets (China’s factory activity just hit its lowest level since March 2009). No doubt Chinese policymakers will strive to lower lending costs to the embattled manufacturing sector as the latter accounts for about one-third of the country’s GDP and employs 15% of all workers. This will be accompanied by a concerted effort to ease China’s manufacturing/industrial overcapacity issues through more infrastructure investments both domestically and in China’s neighboring countries (encouraged by loans through the Asian Infrastructure Investment Bank, for example).

I contend, however, that the PBOC is done with cutting its one-year policy rate for this year, as Chinese policymakers are dealing with a more pressing issue: stabilizing the Chinese currency, the yuan, against the US$ in the midst of recent capital outflows (Goldman Sachs estimates that China’s August capital outflows totaled $178 billion). Simply put–by definition–a country cannot prop up its currency exchange rate while easing monetary policy and maintaining a relatively open capital account at the same time. With the PBOC putting all its resources into defending the yuan while capital outflows continue, it will be self-defeating if the PBOC cuts its policy rate at the same time. The PBOC’s current lack of monetary policy flexibility is the main reason why Chinese policymakers are trying to find ways to stem capital outflows.

Rather than easing monetary policy, Chinese policymakers are utilizing other means to directly increase economic growth, such as: 1) Cutting minimum down payment requirements for first-time home buyers from 30% to 25%, 2) Approving new subway projects in Beijing, Tianjin, and Shenzhen worth a total of $73 billion over the next six years, and 3) Cutting sales taxes on automobile purchases from 10% to 5%, effective to the end of 2016. I expect the PBOC to regain its monetary policy flexibility by early next year, as the combination of record-high trade surpluses and still-low external debt should allow China to renew its policy of accumulating FOREX reserves yet again.

Chinese Casino Gaming Companies Up Despite Chinese Stock Market Rout

Just over two weeks ago, I was interviewed by CNBC Asia; they asked what I was advising my clients to purchase in the Chinese stock market. I specifically mentioned two Chinese companies that are traded offshore – in this case, I discussed Melco Crown (MPEL) and Las Vegas Sands (LVS). MPEL is up +10.1% while LVS is up +6.7% since the day of the June 21st interview. Meanwhile, the Shanghai Composite Index is down by 16.7% in the same time frame.

Policymakers have dealt Chinese casino gaming companies a bad hand (pun intended), e.g. limiting the number of Macau visas, full smoking ban at the casinos, and monitoring Chinese VIP customers, but in general, Chinese casino gaming companies are well-run; despite the slowdown in visitations over the last 18 months, Chinese casinos are enjoying positive cash flow in an oligopolistic market (the government has only awarded six gambling licenses and have limited the number of tables available to patrons). Both MPEL and LVS are poised to take advantage of the ongoing growth in Chinese casino gaming & entertainment spending, driven by the more profitable mass-market clientele in the future.

Leading Indicators Suggest Lower U.S. Treasury Rates

In two of our most recent commentaries (April 10, 2015: “Our Leading Indicators Still Suggest Lower Asset Prices” and March 12, 2015: “The Weakening of the CB Capital Global Diffusion Index Suggests Lower Asset Prices“), we discussed why Goldman Sachs’ Global Leading Indicator was giving highly misleading leading signals on the global economy given its over-reliance on components such as the Baltic Dry Index and commodity prices–both of which could be highly impacted by idiosyncratic factors such as supply disruptions or technological substitutions. Indeed, Goldman itself has been highly transparent and critical over the last six months about the distortions created by an oversupply of dry bulk shipping capacity and an impending wall of additional supply of industrial metals, such as copper and iron ore.

Indeed–because of these distortions–Goldman’s GLI has been highly volatile over the last six months. Last month’s GLI suggested the global economy was “contracting” from January-March 2015–which in retrospect, does not make much sense. Meanwhile, our own studies had suggested that global economic growth was still on par to hit 3.5% in 2015–while our earlier studies suggested U.S. economic growth could hit as much as 3.0%–with energy-importing countries such as India projected to accelerate to as much as 7%-8% GDP growth.

Because again of such idiosyncratic factors, Goldman’s GLI this month suggests the global economy is now moving into “expansion” mode. January data was revised and now suggests the global economy was merely “contracting” that month, with February-March barely in contraction phases. None of these make sense. The latest upbeat data is due to: rising base metals prices, a bounce in the AU$ and the CA$, and a bounce in the highly volatile Baltic Dry Index. Copper’s latest rise was arguably due to Chinese short-covering–Chinese property starts/fixed asset investments remain weak, although we are optimistic that both Chinese commercial and residential inventories are re-balancing.

Our own studies suggest the global economy has been slowing down significantly since the 2nd half of last year; more importantly, the negative momentum has not abated much (despite the re-acceleration of Western European economic growth). Specifically, we utilize a global leading indicator (called the CB Capital Global Diffusion Index, or CBGDI) where we aggregate and equal-weight the OECD leading indicators for 29 major countries, including non-OECD (but globally significant) members such as China, Brazil, Turkey, India, Indonesia, and Russia. The OECD’s Composite Leading Indicators possess a better statistical track record as a leading indicator of global asset prices and economic growth. Instead of relying on the prices of commodities or commodity currencies, the OECD meticulously constructs a Composite Leading Indicator for each country that it monitors by quantifying country-specific components including: 1) housing permits issued, 2) orders & inventory turnover, 3) stock prices, 4) interest rates & interest rate spreads, 5) changes in manufacturing employment, 6) consumer confidence, 7) monetary aggregates, 8) retail sales, 9) industrial & manufacturing production, and 10) passenger car registrations, among others. Each of the OECD’s country-specific leading indicator is fully customized depending on the particular factors driving a country’s economic growth.

The CBGDI has historically led or tracked the MSCI All-Country World Index and WTI crude oil prices since November 1989, when the Berlin Wall fell. Historically, the rate of change (i.e. the 2nd derivative) of the CBGDI has led WTI crude oil prices by three months with an R-squared of 30%, while leading the MSCI All-Country World Index slightly, with an R-squared of over 40% (naturally as stock prices is typically one component of the OECD leading indicators).

Since we last discussed the CBGDI on April 10, the 2nd derivative of the CBGDI has gotten weaker. It also extended its decline below the 1st derivative, which in the past has led to a slowdown or even a major downturn in the global economy, including a downturn in global asset prices. Figure 1 below is a monthly chart showing the year-over-year % change in the CBGDI, along with the rate of change (2nd derivative) of the CBGDI, versus the year-over-year % change in WTI crude oil prices and the MSCI All-Country World Index from January 1994 to May 2015. All four indicators are smoothed on a three-month moving average basis:

OECDleadingindicators

The CBGDI has also led the U.S. 10-year Treasury rate on most occasions over the last 20 years. Whenever the 2nd derivative declines to near the zero line (and continues down), U.S. 10-year Treasury rates have declined 86% of the time over the next 3, 6, and 12 months. Yes, we did enjoy a secular bull market in the U.S. long bond over the last 20 years, but 86% upside frequency is still a very good track record during a secular bull market. The track record is especially attractive considering that: 1) when this indicator was wrong, the worst outcome was a 27 bps rise (over 3 months beginning December 2004); 2) when this indicator was dead on, the best outcome was a highly-profitable, and highly-asymmetric, 168 bps decline in the U.S. long bond (over 12 months beginning December 2007).

10yeartreastudy

As of this writing, the U.S. 10-year rate is trading at 2.18%, which is 14 basis points higher than the average 10-year rate of 2.04% during March 2015, when the 2nd derivative of the CBGDI essentially touched the zero line. As we discussed in past newsletters (and will further elaborate this weekend), we do not believe the ECB has lost control of the Euro Zone’s sovereign bond market. Combined with the ongoing BOJ easing, both central banks are still projected to purchase another $1 trillion of sovereign bonds over the next 12 months. With the U.S. federal budget deficit still near its lowest level over the last six years–and with the People’s Central Bank of China proactively lowering interest rates–I do not believe the U.S. 10-year Treasury rate has any room to move higher from current levels. As such, we are advocating a long position in long-dated U.S. Treasuries; our Absolute Return Liquidity strategy now has a sizable position in the long-dated Treasury ETF, TLT.