Abstract

Current cryptocurrency market is becoming more enriched in the ecosystem by more diversified participants and more influx of funds. In this context, new tools for investments are needed in order to conduct cyclical analysis select from various asset classes across the whole market. Therefore, we have conducted a study on a famous and mature model in traditional finance, the investment clock first proposed by Merrill Lynch, hoping to apply the gist on the crypto market.

The investment clock is a macroeconomic analysis and broad asset allocation model proposed by Merrill Lynch in 2004. It adopts OECD output gap and CPI inflation data as the two main indicators to measure economic growth and price level. By analyzing economic data of the United States since 1970, a series of conclusions were drawn, namely, according to the alternating rise and fall of output gap and CPI data, the economic cycle can be divided into four cycles, namely, reflation, recovery, overheat and stagflation, which occur one after another and constitute a complete cycle. An optimal strategy of asset allocation in each cycle is given according to the derivation and validation of the model.

After the global financial crisis starting 2008, the central banks, represented by the Federal Reserve, have been keeping low interest rate policies for a long time and implemented quantitative easing several times to stimulate economic growth; the investment clock is somewhat affected by the adjustment in the following perspectives: the duration of the reflation and stagflation is shortened, and the duration of the recovery period becomes longer; the clock becomes incomplete that the cycles alternate not following the order. The volatility of various asset classes becomes greater. Stocks perform well throughout the ascending period of economic growth, but the well-performed assets in different phase still fit in the framework of the investment clock.

In the analysis, the position of the crypto market was fitted with the most suitable position in the framework of the investment clock along with which phase in the investment clock best describes current market as well as the best feasible asset allocation. It can be concluded as: we are in reflation; the optimal allocation in current phase is bonds. When it comes to the allocation in equities, defensive stocks and long-term growth stocks are preferred; within the cryptocurrency market, defensive cryptocurrencies, such as BTC and LTC, and long-term growth coins, such as ETH and MATIC, are preferred.

1. The Original Investment Clock

As the cryptocurrency market prospers, the total market capitalization has seen continuous growth enriched by all kinds of projects in increasingly emerging segments. In the past two years, more and more traditional financial institutions and brick-and-mortar companies have entered the cryptocurrency market, investing or participating in project development. A new tool for investment in the market is urgently needed for asset screening and cyclical analysis as the market is becoming a more abundant ecosystem filled with a variety of investors and large influx of funds. Therefore, a study on the famous and mass-adopted the investment clock of Merrill Lynch is conducted with the intention to duplicate it in the crypto market. Since the study is enormous in workload and complex in nature, it will be presented in the form of a series of articles later. This article, as the first one in the series, will first analyze the original investment clock and compare changes when QE is added.

Merrill Lynch first introduced the concept of the investment clock in 2004 in the report, The Investment Clock: Making Money from Macro. The report, through the statistics of the U.S. economic data over the past 30 years, inputting CPI growth YoY and output gap as indicators, divided the economic cycle into four phases: reflation, recovery, overheat and stagflation; the optimal allocation of assets for each stage is also summarized. A more thorough analysis of the model is presented below.

1.1 Analysis

The two parameters in the investment clock to define economic cycles are the CPI growth rate YoY and output gap. The output gap is the gap between actual output and full potential output (i.e., capacity) of an economy. The output gap can be positive or negative: when actual output exceeds capacity, the output gap is positive, which occurs when demand remains high and leads to higher prices; when actual output is below capacity, the output gap is negative, which implies the existence of idle supply due to weak demand and a fall in prices.

Calibration of the output gap is one of the major difficulties in the model: the level of potential output cannot be directly observed; it can only be estimated by various methods. The common method of estimation is to select PMI, GDP, indicators of labor and capital inputs to form a production function, and in a Hodrick Prescott Filter, which could derive long-term trend and short-term fluctuation from the function, an estimate of potential output can be obtained by eliminating the short-term fluctuation. Alternatively, the value could also be received by conducting primary research, such as field surveys of manufacturers. However, none could rule out the possibility of deviation. In addition, the accuracy of the model is highly subject to types of data inputs, methodology to produce the function and estimate, which is the secret of each research institution of investment. The research in this area is so extensive and varied that it could be the subject of a separate article, so no further discussion will be given in this article.

From the chart above, the investment clock essentially classifies the economic cycle by the factors of economic growth and price level, and adjusts investment portfolio by selecting the most suitable asset class based on the yield in a certain phase of economic cycle. The logic behind is that macroeconomic conditions of each country will affect the performance of the various asset classes in the market, prompting adjustments to be made on monetary and fiscal policy, and the adjustments will once again affect the macroeconomic condition as well as various classes. In short, cyclical changes in macroeconomic conditions and economic policies simultaneously lead to cyclical performance of various assets.

According to the framework of the investment clock, an economic cycle can be divided into four phases, and each phase has corresponding outstanding assets.

● Reflation: Economic growth and price level decline together, even negative in GDP growth rate or CPI could be spotted. During this period, the stock market is weak as commodities decline and corporate earnings pale, and the yield curve of bonds shifts down and steepens; interest rate cuts are nornally adopted under such circumstance by central banks in an effort to reboot the economy. Bonds are the best feasible investment in this phase, and defensive stocks are preferred in terms of equity investment. For cryptocurrency, pillar assets in the industry, such as BTC and ETH, are preferred.

● Recovery: Although monetary policy of low interest rate and stimulating economic policies gradually boost the GDP growth back on track, idle capacity is not fully utilized, in other words, the output gap is still negative, resulting in falling in price, thus, a decrease in CPI. Benefiting from loose monetary and fiscal policy during this period, corporate profits start to recover, and stocks become the best investment; cyclical stocks and short-term growth stocks.

● Overheat: As capacity recovers, the output gap gradually disappears, and inflation begins to rise. This is when a central bank starts to raise interest rates to ensure sustainable growth and prevent overheating, but the economic growth remain strong. The yield of bonds begins to fall and becomes flat. The stock market is on a seasaw with increase in corporate earnings on the one side and high valuations on the other, so that cyclical stocks and short-term growth stocks are preferable. Commodities are the best investment option during this period thanks to the soar in inflation.

● Stagflation: GDP growth slows down and productivity declines, constrained by postivie output gap and strong demand, firms raise product price to set off the simultaneous increase of wage and price, hence preserve profits. During this period, tighter monetary policy will be in place until inflation weakens, which depresses the performance of bonds; meanwhile, stocks appear to decline due to tighter monetary policy and lower corporate profits. Cash and cash equivalent assets are the best investment; for stocks, defensive stocks and long-term growth stocks are preferable.

The framework of cycles in the investment clock is also conducive in the research of allocation strategy for stocks and cryptocurrencies:.

● When economic growth accelerates, cyclical industry stocks, such as steel and automobiles are perferred for equity investment, while L1 chain tokens and Defi tokens are preferred for crypto investment; when economic growth slows down, non-cyclical and defensive stocks are relatively better, and BTC is preferred in crypto.

● When inflation is falling, the fiancial costs are relatively low, long-term growth stocks and fundamental compositive projects, such as L1 chains, are preferred; When inflation is rising, commodities and cash or cash equivalent perform the best, therefore, stocks and tokens with potential to be hyped, such as NFT or GameFi tokens, are the best choice.

In summary, the investment clock divides the economic cycle into four phases with economic growth and inflation as the two axes, integrating asset class selection and economic cycle with the model. The inherent rationale is that the two factors, economic growth and inflation, drive the adjustment of national economic policies, and affect the cyclical performance of major asset classes all along. The model will be justified and validated below.

1.2 Data Validation

In this section, annual nominal output gap data published by the Congressional Budget Office (CBO) and CPI data YoY published by the Bureau of Labor Statistics will be fitted into the model for validation; the input data range is extracted from 1970–2007 for the analysis. After 2008, the emergence of new monetary instruments, such as QE, has added variations on the model, which will be elaborated in the next section.

The output gap and CPI data YoY are divided into four cycles according to the investment clock as below. The duration of each cycle and the performance of the major asset classes in each respective cycle are tabulated.

Two typical cycles of the investment clock with four phases each are illustrated above. The duration of each phase is retrieved in sum and in average out of 152 quarters as well as the percentage accounted in the population, as demonstrated in below:

From Table 2, it is evident that in the 37 years from 1970 to 2007, the economic was upward in 63.81% of the time, and 36.19% being downward for the rest; the inflation is increasing for about 54.6% of the time and decreasing for the rest 45.4% of the time. In addition, in terms of the average duration of each cycle, the average duration of the recovery and overheat exceeds that of the stagflation and reflation, with the average duration of the recovery over 1.5 years and the average duration of the overheat over 2 years; while the average duration of the stagflation is roughly over 1 year and almost 1.5 years for reflation, respectively. However, because the duration of each cycle is long and the overall sample size is small, the results are largely biased by outliers. To be more specific, the reflation in 1980–1982 lasted 11 quarters, and the overheat in 1987–1989 and 1994–1996 also lasted 11 quarters; these samples have a large impact on the overall statistics.

In the next step, performance of various asset class in different phases will be classified, where the 10-year U.S. Treasury data are inputted for bonds, the S&P 500 for stocks, the Bloomberg Commodity Index for commodities, and the 3-month U.S. Treasury rate for cash and cash equivalent; performance of each asset class is calculated separately as quarterly average over different phases. The performance of same type of asset but in different time period can differ dramatically from each other as inflation can tremendously vary by time period. For instance, the inflation was 20% at highest in the 80s, but remained under 3% for quite long since entering the 21st century. Therefore, the actual performance is as the followed when eliminating the influence of inflation:

From Table 3, the following characteristics are noteworthy:

●Reflation: 10-year Treasury bond had the best performance, and short-term Treasury, represented by 3-month Treasury, also performed well as the yield of bond steepened again. Correspondingly, commodities were the worst in this phase.

●Recovery: Stocks had the greatest performance as corporate profits were gradually recovered, and bonds became weaker as the capital flow turned to stocks, while commodities were still the worst due to falling product prices and decsending inflation.

●Overheat: As the inflation rate was generally high during this phase and central banks started to raise interest rates, the actual yield of bonds, cash and cash equivalent and stocks were affected and appeared in negative, while commodities were positive and strong.

●Stagflation: As central banks continued to raise interest rates and economic growth declined, leaving bonds, stocks and commodities under pressure with poor perfomance, while cash and cash equivalent outperformed.

After data validation, it is statistically significant that the performance of major asset classes is substantially different in the four phases in the investment clock, and it is consistent with the proposed summary of asset performance in different phases of the investment clock. Therefore, it is reasonable and rational to serve the investment clock as the framework for analysis of macroeconomic cycles and a fundamental strategy for asset allocation.

2. The Investment Clock with QE

In late 2008 and early 2009, the subprime mortgage crisis in the US triggered a global financial crisis. During this period, the output gap and inflation rate of the U.S. plunged simultaneously and both turned negative. In response, the Federal Reserve intervened with Quantitative Easing (QE) as a new tool on monetary policy. Since then, QE has been utilized multiple times to stimulate the economy. What are the influences of QE on the investment clock? Explanations are made in next chapter.

2.1 Introduction of QE

Before further study of QE, one must know commonly adopted methods in monetary policy by the Federal Reserve and how it could affect market liquidity and interest level. In general, traditional monetary policy instruments mainly operate on interest rates. Interest rates can be simply understood as the cost of money. When the economy is overheated and inflation is too high, a central bank raises the cost of financing for businesses and individuals by raising interest rates, thereby discouraging businesses from expanding production and individual consumption; conversely, when the economy is in decline and inflation is falling, the central bank stimulates the economy by cutting interest rates to stimulate business expansion and individual demand.

In United States, for example, the Federal Funds Rate (FFR) is the main instrument to adjust interest rate by the Federal Reserve, which is the market rate for interbank lending in the US, mostly the overnight lending rate. By adjusting this rate, the Federal Reserve can directly affect the cost of funds for commercial banks, which in turn affects the entire macroeconomy. However, it is rather long and intricate for the overnight lending rate, as the starting point, to spread the designated effect to adding costs on every corner of the macroeconomy, not to mention many other factors are beyond control of the Federal Reserve. Therefore, when the financial turmoil happened in 2008, most classic monetary policy instruments partially failed, and QE was engaged by the Federal Reserve as a monetary tool in order to save market liquidity, especially to curb the spike in long-term interest rates.

Different from direct adjustments on FFR, QE is an appearance of the Federal Reserve’s shift of intervention to the supply of funds instead of the cost of funds, aiming to sustain the financial system under an environment with loose liquidity. A commonly adopted method is to purchase treasury bonds and corporate bonds while lowering the FFR; in this case, long-term interest rate is thus lowered and the market is more liquid. This approach in monetary policy was first proposed by the Bank of Japan in 2001 and became famous worldwide when it was adopted by the Federal Reserve during the financial crisis in 2008. From 2008 to present, the following table demonstrates the QE implemented in the United States.

QE is profound to countries all over the world as many economic entities have more or less adopted QE since it was born in Japan and frequently engaged in the US economy. Recall the investment clock by fitting data after 2008 with the intervention of QE, comparisons on economic cycles before and after QE are made and introduced in the next.

2.2 Data Validation with QE

In this section, data inputs are from the same data sources but vary in time range; the new data sets ranges from 1/1/2008 to 3/31/2022. The complete fitting of the model is illustrated below:

Figure 3 demonstrates the investment clock as different economic cycles, and the total and average duration of each phase in the cycle and the percentage account for the whole time are summarized in below:

By comparing the cyclical data before and after 2008, the following characteristics are found:

a) There is not any complete cycle after 2008 following the four-phased economic cycle, but often a reflation was followed by an overheat, or a stagflation interspersed between a recovery and an overheat.

b) During this period, the output gap remained negative for a long time, lasting from late 2008 to early 2019, a duration that has not been seen in history. Meanwhile, the CPI YoY for the same period remained below 3% for a long time, essentially at the lowest level in the selected range. The combination of the two data leads to a conjecture that there was a profound shift in the U.S. economic growth pattern during this period compared to historical.

c) Comparing the data on the average duration and the respective percentage accounted for total time of each cycle over the two time periods, the average duration of stagflation and reflation decreased to 2–3 quarters from more than 1 year prior to 2008, while the average duration of recovery increased by almost 1 year, and the average duration of overheat remained basically unchanged. In terms of percentage of each phase accounted for in total time, stagflation and reflation decreased by about 5.5% and 8%, respectively, while recovery increased by 10% and overheat by about 4%. In sum, compared to the previous period, the average duration and percentage of total time for both stagflation and reflation decreased significantly after 2008, while the average duration and percentage of recovery increased significantly.

The following table summarizes performance of different asset class after 2008 with the same methodology with that before 2008:

Comparing the above results to that prior to 2008:

● The asset class with best performance in each phase matches the investmetn clock: bonds in reflations, stocks in recoveries, commodities in overheats, and cash or cash equivalent in stagflations.

● Changes in bonds and cash or cash equivalents are more significant that that before 2007. For example, the average bond yield during deflations was 14.9%, higher than 3.8% in the same phase before 2007, while cash and cash equivalents during the stagflations rose by 236.6%, far more than 3.2% in the same phase before 2007. It is attributed to that the Federal Reserve had to increase the FFR over 20% in response to the high inflation in the 80s, and it gradually decreased over the next years and maintained at 0.25% for a long time after 2008. In the interest rate hike cycle of 2015 to 2018, however, the Federal Reserve raised the FFR to 2.5%, which is 10 times more than the previous one with rather low basis, and it was not peculiar to see large fluctuations on cash and cash equvalents henceforth.

● Stocks have also achieved higher positive returns during recoveries and overheats unlike before 2008; it coincided with the great bull market in the U.S. stock market that lasted for many years, and it was essentially the outcome of the almost zero FFR and the abundant liquidity.

Through the above analysis, significant changes are effective in the economic growth pattern of the US after the launch of QE after 2008, which are responsible to a prolonged negative output gap and a sticky inflation rate below 3% in the US; never a complete economic cycle described in the investment clock was spotted in the order of reflation, recovery, overheat and stagflation during this period due to the accommodative monetary policy of the Federal Reserve and the repeated intervention of QE. The average duration and percentage of total duration of reflation and stagflation are significantly less than before 2008, whereas the average duration and percentage of total duration of recovery are greater than before 2008. Nonetheless, asset class with best performance is consistent with the conclusions in the investment clock. To sum it up, the long-lasting low interest rate policy and QE after 2008 could be deemed as stimulants that significantly shortened the duration of economic downturn, but it did not reduce the frequency of the downturn and disrupted the complete cycle of the economy.

3. Analysis of current economic cycle and cryptocurrency allocation

According to the investment clock, current phase can be deduced by thorough analysis of current output gap and CPI. Since the annual nominal output gap data published by the Congressional Budget Office (CBO) is currently updated only through the first quarter of 2022, the U.S. output gap data published by the Federal Reserve Bank of St. Louis and the CPI data YoY published by the Bureau of Labor Statistics (BLS) will be inputted for the analysis:

From Figure 4, the output gap in the first two quarters decreased while the CPI increased; the data is in line with the characteristics of stagflation. The CPI has been falling since the beginning of the third quarter, while the output gap is at low level, and according to recent PMI and other economic data, the output gap may fall again in the fourth quarter, on which we could speculate reflation for current period. Furthermore, following the Federal Reserve’s route of interest hikes, it may last until the first quarter of 2023, and it can be inferred that deflation will remain until then. According to the investment clock, inflation decreases and economic growth slows in reflation; the best feasible asset class is bonds, while defensive stocks and long-term growth stocks are most suitable for investment in stock in this period.

In the next, by the classification of stocks, crypto assets are categorized into defensive crypto assets and cyclical crypto assets, where cyclical crypto assets are further divided into long-term growth and short-term growth, as demonstrated in the table below:

Following the framework of the investment clock, defensive tokens, such as BTC and LTC, and long-term growth tokens, such as ETH, UNI and MATIC, are preferred at current phase of reflation.

4. Summary and User Guide of the Investment Clock

The investment clock is a macroeconomic analysis and broad asset allocation model proposed by Merrill Lynch in 2004. It adopts OECD output gap and CPI inflation data as the two main indicators to measure economic growth and price level. By analyzing economic data of the United States since 1970, a series of conclusions were drawn, namely, according to the alternating rise and fall of output gap and CPI data, the economic cycle can be divided into four cycles, namely, reflation, recovery, overheat and stagflation, which occur one after another and constitute a complete cycle. By data validation and

After the global financial crisis starting 2008, the central banks, represented by the Federal Reserve, have been keeping low interest rate policies for a long time and implemented quantitative easing several times to stimulate economic growth; the investment clock is somewhat affected by the adjustment in the following perspectives: the duration of the reflation and stagflation is shortened, and the duration of the recovery period becomes longer; the clock becomes incomplete that the cycles alternate not following the order. The volatility of various asset classes becomes greater. Stocks perform well throughout the ascending period of economic growth, but the well-performed assets in different phase still fit in the framework of the investment clock.

The following points need special attention when the investment clock is applied:

● Source of output gap data. Since the output gap data is inherently an estimate, it may vary by institutions as different methods and data inputs are inserted in the calculation, which may lead to differences when defining phases following the framework of the investment clock.

● Accuracy of output gap data. The output gap data is published quarterly in most institutions; if more frequent and accurate data is desired, custom economic parameters can be selected to construct unique output gap data as long as it is reasonable; the accuracy of data is then subject to the parameters selected and the method of calculation.

● Turning point in the investment clock. Admittedly, accurate and frequent data is able to come to the aid in deciding current economic phase, but finding the turning point is the superior goal to be completed so that asset allocation can be adjusted in time for more profits and less risk exposure.

In previous analysis, the position of the crypto market was fitted with the most suitable position in the framework of the investment clock along with which phase in the investment clock best describes current market as well as the best feasible asset allocation. It can be concluded as: we are in stagflation; the optimal allocation in current phase is bonds. When it comes to the allocation in equities, defensive stocks and long-term growth stocks are preferred; within the cryptocurrency market, defensive cryptocurrencies, such as BTC and LTC, and long-term growth coins, such as ETH and MATIC, are preferred.

Further research will be conducted later on the basis of current research, and more meticulous and quantitative studies will be done on segmentations in the crypto market, incorporating the framework of the investment clock and validating with historical data.