viernes, 2 de agosto de 2013

viernes, agosto 02, 2013

The Risk Of Runaway U.S. Inflation With A Fed In Handcuffs

Jul 31 2013, 09:37

by: Skeptical Investor



This article suggests that the actions by the Fed in recent years to support the economy by reducing its funds rate to the zero bound and greatly expanding its asset portfolio have created conditions of great risk for future runaway inflation in the US. And that this could occur under conditions where the Fed could be seriously constrained in its ability to react. (The "handcuffs" in the title are meant to be metaphorical, not literal.) Since the path of the argument below is not a short one, we begin here with a brief roadmap:

A cursory overview of Fed history is presented to provide perspective on the departure of the recent Fed actions from any policies or methods employed by the Fed in the past. It is then argued that these actions have created a set of economic circumstances that is analogous to the conditions of certain other highly dangerous systems. These are systems that persist in an increasingly unstable equilibrium until some triggering event causes a collapse of significant and perhaps catastrophic magnitude. In the present case, the form of collapse considered would be the onset of uncontrolled inflation. A common element of such systems subject to catastrophe is the presence of a transmission mechanism between the diverse system components that serves to propagate and amplify the triggering event, which itself can be relatively small. In this case, we suggest that the propagation mechanism is the interaction between actual inflation and inflation expectations in the economy.

Finally, we review the available Fed responses to such an inflation event and consider the nature of the restraints on Fed actions that follow directly from the enormous size and the composition of its present and projected asset portfolio. Some brief recommendations are presented in the conclusions for possible investment positioning to provide protection in your portfolio against the risks that we identify.

A recent speech delivered by Fed Chairman Ben Bernanke on the history of Fed operation over the past 100 years since its formation provides the basis for our overview here. The Fed's goal at its formation was to avoid threats to financial stability believed to be periodically produced by shortages of liquid funds in the economy. The Fed was empowered to provide needed liquidity during such episodes though loans to member banks at its discount window, enabling these banks to support the needs of the markets.

The result was that the Fed tended to increase the money supply during periods when the economy was most active and upward pressures on prices was highest. This "pro-cyclical" bias and the resulting unstable behavior of both the money supply and inflation is very evident in our Figure 1, which shows the history of the monetary base maintained by the Fed since its inception and the CPI over the same period. (The monetary base data here, as well as other monetary data presented further below is taken from tables maintained by the St. Louis Fed, while the CPI data is that provided by the Bureau of Labor Statistics.) This lack of inflation control by the Fed clearly persisted through the great depression and only began to change after the end of World War II.

(click to enlarge)
Figure 1. A long-term history of changes in the Fed's monetary base and CPI inflation, and more recent Fed economic action exerted through the funds rate.


By this time, the effects of the great depression and the return of soldiers from the front had raised the perceived importance of encouraging full employment as a goal of Fed activity. Congressional actions starting in 1935 increased the independence of the Fed, expanded its mandate well beyond that of simply providing needed liquidity, and allowed a greater scope of available monetary actions. In the decades after the war, the Fed began to use open market operations (originally intended only as a source of self-funding for the Fed) and the discount rate to influence short-term interest rates, with the Fed funds rate target eventually becoming the dominant economic control parameter. However, the need for the use of these new tools to actively control inflation was not sufficiently appreciated until after the two bouts of high inflation experienced in the 1970's. Under Paul Volcker's leadership, starting in 1979, the Fed became more willing to respond with its already existing tools to fight high inflation, using the massive increase in the Fed funds rate seen in Figure 1 to hammer down the inflation threat.
 
The evident success of Fed policies during the following decades was attributed to the Fed's new willingness to respond aggressively with monetary policy both to threats to financial stability such as the 1987 market crash, and to inflationary pressures. It was also increasingly accepted that the Fed's ever-increasing use of communications to promote public understanding of the Fed's goals and actions was an important ingredient of its success. Most particularly, "providing a firm anchor, secured by the credibility of the central bank, for the private sector's inflation expectations" is cited by Bernanke as a key factor in the Fed's success. This point of view was also greatly emphasized in a speech given in 2007 by Fred Mishkin while he was a member of the Fed's Board of Governors (available in an NBER working paper). There, he strongly credits the Fed's control of long-term inflation expectations for the highly desirable changes in inflation dynamics observed during the preceding decades, including the reduced persistence of inflation, with only a transitory response to economic shocks.
 
Mishkin also briefly discusses the near-term impact of inflation expectations, which he states have become a "central feature of the increasingly popular New Keynesian Phillips curves, in which current-period inflation is a function of expectations of next period's inflation and resource utilization." An introduction to the corresponding approach to modeling inflation dynamics may be found in a 2008 paper by Andreas Hornstein (at the time a vice president at the Richmond Fed).

The conclusion that near-term inflation expectations drive inflation is not surprising, in that it simply reflects the natural responses of "economic actors" (e.g., buyers, sellers, borrowers and lenders) to anticipated inflation changes. For example, if inflation is expected to increase, buyers will tend to accelerate purchases, while sellers will attempt to delay sales out of inventory, both of which actions would increase shortages of desired items, leading immediately to higher pricing pressures. Below, we will discuss and demonstrate the reverse near-term relationship, indicating that the level of inflation experienced also drives near-term inflation expectations.
We believe that the existence of this two-way dependence between inflation and its expectations is important, since this is identified as a possible transmission mechanism in the production of runaway inflation in the US economy under its currently unstable conditions.
 
Remarkably, we see in Figure 1 that moderate CPI inflation has persisted through the recent years since the onset in the summer of 2007 of financial crisis, the subsequent "great recession", and the aftereffects that still continue. The immediate Fed response to the financial crisis was to quickly lower the funds rate to its lower bound near 0%.

Subsequently, the monetary base was raised through a series of asset purchase programs to support financial stability and to prevent an economic spiral into deflation. The large rapid fluctuations in the rate of change of the monetary base during this period reflect the onset and completion of these various programs. In Figure 2, we show the corresponding history of the monetary base, bank reserves held at the Fed, the M1 money supply, and the nominal GDP during the period from the late 1950's through the present. Note that the vertical scale is logarithmic, which tends to make the increase in the monetary base since the onset of the financial crisis look smaller than its actual amount, about $2.35T (trillion). Just the initial abrupt increase in the monetary base at the start of this process was about $1.27T. A monetarist would assert that massive inflation has already occurred as a result of this degree of money creation. It's founder, Milton Friedman, famously stated that "Inflation is always and everywhere a monetary phenomenon". In this point of view, the appearance of this inflation in other measures such as the CPI has only been delayed, not avoided.

(click to enlarge)
Figure 2. History of the monetary base, bank reserves held at the Fed, the M1 money supply, and the nominal GDP during the period from the late 1950's through the present.


Figure 2 reveals that certain long-term relationships between the quantities shown have been seriously disrupted by the Fed's recent asset purchases, which we will refer to here as their quantitative easing (QE) programs, although not all of them were called by this name. Prior to QE, we find that M1 and bank reserves proceed along approximately parallel paths on the chart. Since the vertical scale of the chart is logarithmic, such parallel paths indicate a constant ratio between the underlying values. In this case, the average ratio of the two quantities over the period shown on the chart before QE is found to be about 14, which reflects the level of money expansion produced largely by bank lending during the period. While QE is taking place, reserves greatly expand but the expansion does not produce a corresponding increase in M1, since there is not enough loan demand from qualified borrowers to absorb the potentially available bank money.

Similarly, prior to QE the monetary base and GDP follow approximately parallel paths, with an average ratio of about 18 (the corresponding money velocity). Again here, the nominal GDP does not follow the monetary base increases produced by QE. If it did, the nominal GDP would have increased by a factor of about 3.15 over the QE period, corresponding to the increase in the monetary base, instead of its actual increase of about 1.11. This would have been an additional increase in nominal GDP by a factor of 2.84, or 184%. Since real GDP can only increase over time as new capacity is brought on line and necessary labor is hired and trained, a short-term increase in nominal GDP would be essentially all inflation, and would be reflected in the CPI. So we take the above result as an indication of the possible magnitude of inflation that could be produced (or has been produced according to a monetarist viewpoint) by the increase in the monetary base. This is an alarming number, which would suggest possible runaway inflation in our future, unless the inflation could be somehow prevented by Fed action.

Certainly, there are at present no obvious reasons why inflation would be expected to increase significantly any time soon, since there is enough excess production capacity and labor available in the economy to avoid any shortages that would exert pressure on prices. And the prices of most raw materials and of energy have been well behaved lately, so there would not appear to be a concern there either. It would seem that it would take some sort of catastrophe for a runaway inflation event to occur under these conditions. But unfortunately we suggest here that there is an important underlying similarity between the current economic conditions and a number of totally different circumstances in which catastrophes are at times produced. We refer here to conditions resulting in such events as earthquakes, avalanches, landslides, and forest fires.

These are examples of phenomena that occur in systems in a state of highly unstable equilibrium, where a change in the underlying conditions can initiate an event of unpredictable magnitude at any time. Moreover, there is usually at most a weak relationship between the size of an initiating cause and the magnitude of the ultimate effect. Such systems have been studied with experimental and computer simulation methods and to some extent modeled mathematically. A particularly accessible treatment of the subject can be found in the book "Ubiquity, Why Catastrophes Happen" by Mark Buchanan (Three Rivers Press, 1999).

To understand the nature of such systems, a good place to start is to consider the production of sand pile avalanches, phenomena that were extensively studied by computer simulation by Bak et al, and documented in a 1987 paper in Physical Review Letters. In the computer simulations, grains of sand are dropped one-by-one at random onto a sand pile, and eventually build up to a critical slope at various locations in the pile, so that if another grain of sand lands at one of these locations, a slide will occur. The sliding grains of sand can then in some cases land on other critical locations, causing further slides to occur. Depending on the previous history of the sand pile, the response to a single new grain of sand can be an avalanche of essentially any size, from a few grains to a catastrophic collapse of the entire pile. In this case, the trivial initial disturbance is propagated by the motion of the sand grains themselves. All such phenomena require some means of propagation of disturbances from one part of the system to other parts.

Another interesting example of such phenomena is the occurrence of forest fires in an area where no fires have occurred for some time, allowing the buildup of tinder at the base of the trees. A source of ignition at a random point within the forest can then spread over a more or less extended area, depending on where the ignition occurs and on the distribution of accumulated tinder and additional flammable materials. Where winds are present, they propagate embers across gaps of various sizes, depending on the wind speeds. We suggest that this forest fire example provides a particularly useful analogy to the present economic situation. The large increase in the monetary base and its potential distribution and amplification by means of lending throughout the banking system is analogous to the buildup of tinder in the forest, and constitutes a system in a highly unstable state of equilibrium. In this view, an inflation shock to the economy of a relatively small magnitude can serve as the "source of ignition", producing an inflation response "forest fire" of unpredictable size at any time. We further suggest that the necessary propagation mechanism of the inflation "disturbance" in this case is the short-term interaction between the inflation effects and inflation expectations communicated among the marketplace participants.

In our historical review above, we cited the near-term dependence of inflation on recent inflation expectations, as embodied in the formulation of New Keynesian Phillips curves. But it is also true that there is a near-term dependence of inflation expectations on recent inflation indications. We found that we could readily show convincing evidence of this latter functionality using the improved inflation expectation data developed recently by Haubrich et al at the Cleveland Fed. The methodology used there to produce the new results is unique in its use of data on inflation swaps, nominal interest rates, and two different expectations surveys.

Historical data for TIP spreads, largely the foundation of earlier methodologies, are not used at all here. The referenced results have been found to provide more accurate assessments of the expectations and are now routinely produced monthly for all forward periods (in years) between one and 30 years. These have been made available for start dates going back to 1982. We used this data in the following relatively straightforward analysis to demonstrate the dependence of near-term expectations on the level of inflation experienced in the economy.

The starting point for our analysis was the monthly CPI-U data from the Bureau of Labor Statistics and the monthly data on 1-year inflation expectations from the Cleveland Fed, which are shown in Figure 3. Note that the monthly inflation data is accumulated over 1-year periods and the result is shown on the figure at the starting month, while the expectation data represent the average expected inflation over the year following the starting month shown on the figure. Thus, if the inflation expectations for the year ahead were perfectly accurate, the two curves would coincide. Clearly this is not the case, implying that the marketplace does not do a very good job of predicting average inflation for the year ahead.

(click to enlarge)
Figure 3. Monthly CPI-U data from the Bureau of Labor Statistics and monthly data on 1-year inflation expectations from the Cleveland Fed.


But still, it seemed as though the curve for the expected inflation has some recognizable features similar to those in the CPI-U curve, although delayed in time by about a year. So the next step was to remove the short-term trends of both curves, which was done by computing the 4-year moving average of each of the curves and subtracting that average from the curve itself to get a monthly residual for each curve. Then, the cross-correlation was computed for the two residual curves as a function of the relative advance (i.e., to an earlier time) in months of the inflation expectation curve, producing the results shown in Figure 4. With no advance, the cross-correlation is negligible, confirming that the marketplace has essentially no ability to predict expected inflation for the year ahead. But as the advance parameter in Figure 4 approaches 1 year, the cross-correlation reaches a maximum.

When the expectation data are shifted in time to 1 year earlier, the correlation is essentially matching the cumulative inflation for each yearly period with the expected average inflation for the subsequent 12 months. Thus, the market's "prediction" in this case for the next 12 month's average inflation is based on what has already occurred during the last 12 months. It is not surprising that leads to the higher cross-correlations seen in the results.

It seems to be the general nature of the marketplace to predict that the future trend in a market parameter will simply continue the recent trend observed, and this is just another case of the same. The peak cross-correlation actually occurs in this analysis at a 10-month advance, and Figure 5 shows the two residual curves overlaid with the inflation expectation residual curve advanced in time by 10 months. The similarity of the two curves is now very evident.


(click to enlarge)
Figure 4. Cross-correlation of the residuals of the accumulated 1-year CPI-U and the expected average 1-year inflation, both relative to their 4-year moving averages.

(click to enlarge)
Figure 5. Residual curves for the annual CPI-U change and expected average annual inflation overlaid, with the expected inflation curve advanced in time by 10 months.


Although the above analysis was limited by the available data to show the dependence of inflation expectations on actual inflation with about a one year delay, we expect that the actual functional relationship also operates over much shorter time frames. After all, the inflation expectations in the marketplace are just an aggregate of the individual opinions (largely emotional) of the population of businesses and consumers (i.e., the actors) in the economy. In fact, in the case of an inflation shock that becomes broadly recognized in the marketplace, we would anticipate the impact on expected inflation to be essentially instantaneous.

Similarly, when inflation expectations change, economic actors can be imagined to almost immediately consider how the change might impact their needs and opportunities, and to quickly act on these potential impacts in order to minimize the damages or take advantage of the opportunities. Cleveland Fed Vice President Joseph G. Haubrich stated the view this way:

People's expectation of inflation enters into nearly every economic decision they make. It enters into large decisions: whether they can afford a mortgage payment on a new house, whether they strike for higher wages, how they invest their retirement funds. It also enters into the smaller decisions, that, in the aggregate, affect the entire economy: whether they wait for the milk to go on sale or buy it before the price goes up.

These considerations lead us to the suggestion that this potentially rapid interaction between inflation and inflation expectations, widely communicated through the multiple information transmission mechanisms used today, can play the part of the propagation mechanism of inflation shock leading to a possibly much larger inflation response throughout the current economy.

Returning to our forest fire analogy, recall that we suggested that the massive increase in the Fed's monetary base and its potential distribution by means of lending through the banking system is analogous to the buildup of tinder in the forest. The rapidly communicated interaction of inflation and inflation expectations plays the part of the wind-blown embers. And some sort of initial inflation shock plays the part of the ignition source.

We also suggested above, based on the comparison between the recent histories of the monetary base and nominal GDP that an inflation shock might result in runaway inflation with a cumulative impact approaching 200%. We can imagine the initial inflation shock to result from any number of plausible events, for example a major natural or man-made disruption to the world's oil supply. The reader is invited to conceive of their own nightmare scenarios offered by current world circumstances. In this analogy, the Fed must play the part of the fire suppression system deployed in the forest. Through the use of its available tools, the Fed would need to confront the inflation process before it goes too far.
Moreover, the Fed's actions must be forceful enough to maintain its credibility since, as the Fed has clearly recognized, only then can its commitment to its long-range inflation target be effective in influencing inflation expectations in the marketplace.

The available inflation-fighting tools of the Fed have changed to provide a better match to their anticipated needs under the present conditions, which are very different from those of the past. Under ordinary circumstances, the Fed would respond to inflation risks by raising the Fed funds rate through open market operations (OMO). More generally, the Fed would adjust the supply of reserve balances to keep the federal funds rate around the target rate established by the Federal Open Market Committee (FOMC). However, these ordinary methods of controlling short-term interest rates in the marketplace would be ineffective while the reserve balances at the Fed remain at the elevated levels resulting from the various QE programs. In order to be able to influence short-term rates, the Fed must somehow first eliminate these excess reserve balances or in effect tie them up in some way so that their OMO can regain effectiveness. These new tools are discussed on the Fed's web site.

Congress, starting in October 2008, granted the Fed the authority to pay interest on their reserve balances. The Fed considers that this will be an especially important method of influencing short-term interest rates, in much the same way as would ordinarily be accomplished through control of the Fed funds rate. Specifically, increasing the interest rate on reserves would be expected to put significant upward pressure on all short-term interest rates, since this risk-free return would compete with the potential income from loans the banks could provide to the marketplace.

Currently, the interest rate paid on reserves by the Fed is only 0.25%, and with the level of these reserves at about $2.1T, the interest expense is approximately $5.25B, which is relatively insignificant. However, if the Fed needed to respond to an inflation risk, the interest paid on reserves would need to be raised to levels similar to Fed funds rates used in the past under comparable circumstances. Figure 6 provides some historical perspective on funds rates that have been used in this way previously. Optimistically ignoring the very high rates used by the Volcker Fed to fight the inflation generated in the 1970's, we see that rates as high as 5% have been used under what might be considered to be more ordinary conditions. If such conditions occurred now, and the interest on reserves were raised to 5%, the annual interest paid on the same reserves would increase to $105B.

This would probably not escape attention, and since it would be paid to banks on their risk-free balances at the Fed instead of being lent in the marketplace, it might be considered quite offensive by the public and by Congress. It could easily be viewed as yet another example of favoritism for "Wall Street over Main Street." And obviously, both the level of reserves and the required interest rate paid could both be higher than assumed here at the time of some future inflation risk, making the situation that much worse.

(click to enlarge)
Figure 6. Until the recent period of near-zero funds rates and massive QE, the funds rate was an important tool in implementing Fed policy.


If this approach was considered objectionable, there are other methods that the Fed could use in an attempt to restore OMO to more normal effectiveness. However, selling the accumulated QE assets would probably not be one of these. Under conditions of an inflation threat, long-term interest rates would be expected to increase along with short-term rates, reducing the market value of the QE assets held in the Fed's portfolio, and increasing the corresponding losses that would be experienced should these assets be sold. To get a rough estimate of the possible losses, we used the Fed data release listing the composition of their portfolio assets for a recent date (July 18, 2013). We found that the major components were about $1.962T of treasuries and $1.235T of MBS, for a total of $3.196T of these assets. The release also provides a breakdown of the assets by maturity, but only in rather broad maturity categories (e.g., 1 to 5 years, 5 to 10 years, and over 10 years). Nevertheless, making simplifying assumptions we used the data to roughly estimate the average portfolio maturity to be about 13 years. Since this is already admittedly imprecise, we just assumed that the average portfolio duration is 10 years, so that a 1% increase in interest rates would result in a mark-to-market loss of about $320B on the portfolio. If the increases in long-term rates (relative to when the Fed's assets were purchased) were about the same 5% assumed above for the increase in short-term rates, the loss would be about $1.6T, a truly daunting result should the assets be sold.

Since the Fed does not use the same accounting methods used by other banks, the losses contemplated above would not actually be recognized as such by the Fed. Historically, the Fed has remitted all net income, after its operating expenses and other items, to the U.S. Treasury. As those net earnings accrue, they are recorded on the Federal Reserve's balance sheet as "Interest on Federal Reserve notes due to U.S. Treasury." In recent years, the Fed income produced by its large and relatively long-term asset portfolio has been substantial. As a result, for example, the remittance to the Treasury in 2012 was a record $88.9 billion, which of course directly reduced the annual deficit. In the event that the Fed net income declined so that it was only enough to cover its expenses, this balance sheet item would fall to zero and the payment to the Treasury would be suspended.

Finally, if there was a net operating loss in some period, then no remittance would be made until earnings, through time, were sufficient to cover that loss. The value of the earnings that would need to be retained to cover such a loss is called a "deferred asset" and is booked as a negative liability on the Federal Reserve's balance sheet under the same line item "Interest on Federal Reserve notes due to the U.S. Treasury." Thus, any losses on Fed's asset sales would be considered a "deferred asset" rather than a loss, and would just appear in its balance sheet in the corresponding line item.

Congress has already noticed the "uniqueness" of such an accounting treatment of losses produced by possible Fed asset sales, has begun to ask questions about it, and could well consider such losses to effectively increase the deficit when the sales take place. This might cause an outrage even greater than would be produced by the annual interest payments on reserves discussed above. 

Furthermore, the above analysis ignores the fact that, to be effective in promptly restoring OMO control of short-term rates, the asset sales would need to take place over a short time frame. This would undoubtedly adversely impact the prices on the assets to be sold, making the situation worse. When Fed Chairman Bernanke announced in his most recent Congressional testimony a contingent plan to begin to reduce the pace of QE purchases later this year and possibly terminate them sometime next year, the interest rate on the 10 year treasury bond promptly increased almost immediately by about 1%, even though not a single bond or MBS had been sold. One can only wonder what the effect of actually reducing the pace of purchases would be on long-term rates, or the effect of stopping purchases completely, not to mention actually quickly selling the accumulated assets.

It thus seems highly doubtful that such asset sales will take place in the context of some type of inflationary threat that would require a strong response by the Fed. Some alternate responses are discussed on the Fed's website:

The Federal Reserve has been developing other new tools to drain large quantities of reserves, including reverse repurchase agreements (reverse repos) and term deposits. Reverse repos are transactions in which the Federal Reserve sells a security to a counterparty with an agreement to repurchase the security at some date in the future. The counterparty's payment to the Federal Reserve has the effect of draining reserves. As a second means of reducing reserves, the Federal Reserve has plans to offer term deposits to depository institutions. Funds in such term deposits would not be available to satisfy reserve requirements or clear payments and, as a result, would not count as reserve balances. Together, reverse repos and the term deposit facility would allow the Federal Reserve to drain substantial volumes of reserves from the banking system.

Unfortunately, the effective interest rates on the repos or term deposits would still need to compete with interest rates available for loans in the open market, so the costs to the Fed would probably be similar to that incurred by using the simpler interest rate payments on reserves discussed above. Although these costs are not quite as transparent as the interest rate payments, one suspects that the reactions of the public and Congress would ultimately not be much different. And neither method permanently removes the long-term QE assets from the Fed's balance sheet, so the impact on reserves would only be temporary.

The current federal reserve under Bernanke has taken bold and innovative actions to rescue the economy from a great initial financial shock and the subsequent threat of descent into possible depression. They deserve great praise for their deep insight into the economic fundamentals at play and the creativity displayed in their responses. Although they frequently speak about the risks of their actions to be weighed against the benefits, it is doubtful that they have considered the possibility of the economy ultimately behaving like the highly unstable systems that we have offered for comparison. In the event, it would take considerable fortitude for a future Fed to take the necessary forceful and consequential actions to counter an inflation threat, since the adverse reactions that would likely result could very well threaten their independence or even lead Congress to replace the present system with a new one under their more direct control.

Fears of such end results would constitute the restraints on the Fed implied in the title of this article. But a timid Fed response or interference with a bolder Fed response by Congress to an initial inflation threat would leave the economy open to a possible catastrophe of the sort that highly unstable systems can produce when their equilibrium is disturbed. When the economy eventually finds its new state of equilibrium, this may reflect extremely high levels of accumulated inflation such as were suggested above. We must hope that we are very lucky or that the Fed and Congress find a way to reduce the present instability of the economy before such a runaway inflation is ignited.

But to be prudent, we also suggest that you protect your portfolio against the future threat of major inflation that has been identified here. Recall that we have emphasized the general lack of predictability for the timing of the collapse of highly unstable systems such as we have compared with the current economy. Although the Fed has begun to hint at slowing down the rate of QE asset purchases, it remains to be seen how soon even this relatively minor measure takes place.

Meanwhile the Fed assets continue to accumulate, and the instability of the economy continues to grow. So your portfolio responses must be viewed as long-term positions to be maintained until, one way or another, the economic situation changes. We recommend that positions in bonds and bond-like stocks be reduced or eliminated, and that positions in large, stable companies that have valuations tied to hard assets be accumulated. These should include companies in basic materials industries such as mining (both precious metal and industrial materials), energy, and agriculture, including the companies that support these operations. Consider emphasizing companies headquartered and with assets mostly located in more stable countries that are less likely to resort to measures such as nationalization in times of great stress. Companies such as are recommended here can be very volatile investments in many cases, so you should consider your ability to tolerate this volatility over a possibly prolonged duration.

0 comments:

Publicar un comentario