Statistics and a Centrally Planned Economy
By John R. Taylor, Jr.
Living in the country with the most fabulously disorganized government in the world, with three famous branches and dozens of others which are hard to remember – all of which are attempting to expand their power and their budget by putting their best face forward – we hear a cacophony of information. These many facets of the government power structure deliver statistics to the rest of us and, very often, their views are different, reported with a different angle because they are beholden to different political ideals or different lobbying groups. Added to these official statistics are those of diverse unofficial sources – the Conference Board, Markit, or Bloomberg – who slice and dice reality in their own ways. The competition among data sources means that errors are not tolerated, data can be trusted, and the detail is impressive. As a result, analysts not only have a wealth of data to review, but they explicitly believe it. This plethora of information and the trust surrounding it are all a function of the decentralized nature of the US system. Everybody is checking on everybody else, and no one is boss. We would propose a rule: The quality of data is positively correlated with the decentralization of the control system. If we examine the historical record involving large American, European, and Japanese companies, which, of course, have much more centralized control than the US government, we soon see that the statistics they report about themselves are less trustworthy. In some cases, the deceptive numbers are so pervasive that the companies fail once the real numbers rear their ugly heads. Olympus, Enron, and Washington Mutual come to mind, and the results involved losses of millions of dollars and many, many lawsuits. The centralized nature of business organizations is the reason we have audits, the SEC, and the FDIC. The truth is positively related to competition.
As competition is a critical component of high-quality statistics, we should always take the numbers put forward by command organizations with a large dose of skepticism. Forty years ago, if you were a middle manager in a corporate behemoth like IBM, you were given sales targets, profit targets, and expense targets that you were expected to meet or exceed. Failure was harmful both for your current pay and your future career. As a result, you would do everything possible to fulfill as many of your goals as possible, even if it meant hurting the company in the longer term or twisting the truth a bit. The penalties were not life or death, however, as this was America, and there was another job somewhere else that involved a different power structure – a different ladder to the top. Transfer this IBM example to a system where all the power structures were tightly interrelated, and there really was only one ladder to take. A centralized economic and social system, unified around a political party that involves itself in every step of the economy has much more power than your boss at IBM could ever have had, as there is no GE or Dupont down the street. When the Central Committee says that growth will be around 7.5% in the next year, that order flows down through the entire system, telling every manager in the State Owned Enterprises (SOEs), which dominate the economy, what part they are to play in making that goal. When the local collector of statistical information calls up, and things aren’t going too well, what are they going to say? The odds are each one will be more positive than is warranted. Multiplied millions of times throughout the system, reality is distorted. There is no way that it cannot be. Because of this, China watchers distrust almost all statistics and believe only those that avoid centralized collection methods or ones that are outside of the country. Electricity usage was one of those but, as it becomes popular as an indicator, it threatens to fall increasingly under centralized management and will lose its validity. Recently the official PMI numbers, which have shown inconsistencies in the past, were partly withdrawn from the market as they could not be verified with other sources of data. Although the numbers are looking good today, it is hard to know whether China is actually growing at the rate claimed by the statistics.
Tarred with the Same Brush
By Jonathan Clark
The Canadian dollar has been in a downtrend for nearly a year and it is unlikely it is over yet. Perhaps the most remarkable aspect of the currency is that the normal quantitative relationships have broken down and are showing no signs of reasserting themselves. The most important variable in the value of a currency is capital flows, which are largely influenced by short-term interest rate differentials. This has been a non-factor due to the directionless nature of interest rates in the US versus Canada (green line). At his first meeting the new Bank of Canada, Governor Stephen Poloz tied future hikes in official interest rates (currently at 1%) to economic growth, but without specific targets. The Canadian economy is growing, but at a slower annual pace than the US. Inflation at 1.3% in July is well within the 1-3% BoC target band. Canada is a major exporter of energy but, despite higher oil prices since the middle April, the Canadian dollar is lower since that time. The prospects of a US strike on Syria could cause a spike higher in the CAD due to the normally positive correlation with the price of oil, but it appears unlikely that the impact would last in the current environment. When considering all of the commodities Canada exports on a weighted basis, the value is roughly unchanged from the start of the year. The currency tends to trade with an equity index like the S&P 500, and yet thus far this year the correlation is -79%. Canada is mainly
suffering because of the end of the mining boom and the slowing of capital into this industry. The CAD hasn’t suffered nearly as much as AUD, BRL or ZAR, but it is being tarred with the same brush and this is keeping it under pressure.
USD/CAD should weaken into Friday when a minor low is due and should test the support at 1.0450. It should then recover for a few days before making a further decline into the week of September 23 when a medium-term low is due. Our target for this downmove is the 1.0375 area. Although there is a chance it could fall to as low as 1.0260 before bottoming, this negative outlook is far less likely. Following this low expected in late September, USD/CAD should turn higher and rally into late October when a significant peak is due. If the resistance at 1.0550 breaks at any time then it can trade to as high as the 1.0875 area before peaking, but this is becoming less likely and we expect a continuation of the slow and steady upmove. It now appears that any final strength seen in late October will be modest and then dollar/Canada will turn lower and decline for a minimum of several months.
By John R. Taylor, Jr.
Living in the country with the most fabulously disorganized government in the world, with three famous branches and dozens of others which are hard to remember – all of which are attempting to expand their power and their budget by putting their best face forward – we hear a cacophony of information. These many facets of the government power structure deliver statistics to the rest of us and, very often, their views are different, reported with a different angle because they are beholden to different political ideals or different lobbying groups. Added to these official statistics are those of diverse unofficial sources – the Conference Board, Markit, or Bloomberg – who slice and dice reality in their own ways. The competition among data sources means that errors are not tolerated, data can be trusted, and the detail is impressive. As a result, analysts not only have a wealth of data to review, but they explicitly believe it. This plethora of information and the trust surrounding it are all a function of the decentralized nature of the US system. Everybody is checking on everybody else, and no one is boss. We would propose a rule: The quality of data is positively correlated with the decentralization of the control system. If we examine the historical record involving large American, European, and Japanese companies, which, of course, have much more centralized control than the US government, we soon see that the statistics they report about themselves are less trustworthy. In some cases, the deceptive numbers are so pervasive that the companies fail once the real numbers rear their ugly heads. Olympus, Enron, and Washington Mutual come to mind, and the results involved losses of millions of dollars and many, many lawsuits. The centralized nature of business organizations is the reason we have audits, the SEC, and the FDIC. The truth is positively related to competition.
As competition is a critical component of high-quality statistics, we should always take the numbers put forward by command organizations with a large dose of skepticism. Forty years ago, if you were a middle manager in a corporate behemoth like IBM, you were given sales targets, profit targets, and expense targets that you were expected to meet or exceed. Failure was harmful both for your current pay and your future career. As a result, you would do everything possible to fulfill as many of your goals as possible, even if it meant hurting the company in the longer term or twisting the truth a bit. The penalties were not life or death, however, as this was America, and there was another job somewhere else that involved a different power structure – a different ladder to the top. Transfer this IBM example to a system where all the power structures were tightly interrelated, and there really was only one ladder to take. A centralized economic and social system, unified around a political party that involves itself in every step of the economy has much more power than your boss at IBM could ever have had, as there is no GE or Dupont down the street. When the Central Committee says that growth will be around 7.5% in the next year, that order flows down through the entire system, telling every manager in the State Owned Enterprises (SOEs), which dominate the economy, what part they are to play in making that goal. When the local collector of statistical information calls up, and things aren’t going too well, what are they going to say? The odds are each one will be more positive than is warranted. Multiplied millions of times throughout the system, reality is distorted. There is no way that it cannot be. Because of this, China watchers distrust almost all statistics and believe only those that avoid centralized collection methods or ones that are outside of the country. Electricity usage was one of those but, as it becomes popular as an indicator, it threatens to fall increasingly under centralized management and will lose its validity. Recently the official PMI numbers, which have shown inconsistencies in the past, were partly withdrawn from the market as they could not be verified with other sources of data. Although the numbers are looking good today, it is hard to know whether China is actually growing at the rate claimed by the statistics.
Tarred with the Same Brush
By Jonathan Clark
The Canadian dollar has been in a downtrend for nearly a year and it is unlikely it is over yet. Perhaps the most remarkable aspect of the currency is that the normal quantitative relationships have broken down and are showing no signs of reasserting themselves. The most important variable in the value of a currency is capital flows, which are largely influenced by short-term interest rate differentials. This has been a non-factor due to the directionless nature of interest rates in the US versus Canada (green line). At his first meeting the new Bank of Canada, Governor Stephen Poloz tied future hikes in official interest rates (currently at 1%) to economic growth, but without specific targets. The Canadian economy is growing, but at a slower annual pace than the US. Inflation at 1.3% in July is well within the 1-3% BoC target band. Canada is a major exporter of energy but, despite higher oil prices since the middle April, the Canadian dollar is lower since that time. The prospects of a US strike on Syria could cause a spike higher in the CAD due to the normally positive correlation with the price of oil, but it appears unlikely that the impact would last in the current environment. When considering all of the commodities Canada exports on a weighted basis, the value is roughly unchanged from the start of the year. The currency tends to trade with an equity index like the S&P 500, and yet thus far this year the correlation is -79%. Canada is mainly
suffering because of the end of the mining boom and the slowing of capital into this industry. The CAD hasn’t suffered nearly as much as AUD, BRL or ZAR, but it is being tarred with the same brush and this is keeping it under pressure.
USD/CAD should weaken into Friday when a minor low is due and should test the support at 1.0450. It should then recover for a few days before making a further decline into the week of September 23 when a medium-term low is due. Our target for this downmove is the 1.0375 area. Although there is a chance it could fall to as low as 1.0260 before bottoming, this negative outlook is far less likely. Following this low expected in late September, USD/CAD should turn higher and rally into late October when a significant peak is due. If the resistance at 1.0550 breaks at any time then it can trade to as high as the 1.0875 area before peaking, but this is becoming less likely and we expect a continuation of the slow and steady upmove. It now appears that any final strength seen in late October will be modest and then dollar/Canada will turn lower and decline for a minimum of several months.
No comments:
Post a Comment