Venezuela Projected To Begin 2018 With Under $5.3 Billion in Forex Reserves

Using a linear forecast of Venezuela foreign exchange reserves, and accounting for roughly $3.8 billion in debt payments through the government and state-owned PDVSA through December 2017, I am projecting the level of foreign exchange reserves to end up at just under the $5.3 billion mark on 1/1/2018.

Data source: Central Bank of Venezuela
Chart source: Author's creation

As of July 22, 2017, Venezuela's foreign exchange reserves stood at $9.971 billion, the lowest level of reserves seen during the current crisis. As the chart above shows, the country's foreign exchange reserves have been broadly falling since the start of 2017, from a few ticks under $11 billion at the beginning of the year to the $9.971 billion level we are at now.

I took these two data points and, to satisfy my own curiosity, used the data points to create a linear projection of foreign exchange reserves for each day through the end of 2017. Using these data points, the linear equation was calculated to be:

Y = -7.75x + 10,995

where Y is the level of foreign exchange reserves in millions of USD, and X is the date in basic number format. For example, July 21, 2017 was assigned the number 134, as it was the 134th data point in the foreign exchange reserves data set as provided by the Central Bank of Venezuela. I have filtered out all weekend dates, so the linear projection does not create an artificially-low ending forex reserve level by incorrectly lowering reserves on weekends. It should be noted that this model does not account for holidays or other days off during the workweek.

January 1st, 2018 was the 250th data point in this data set, so the projection was calculated to be Y = (-7.75*250) + 10,995. This resulted in $9.055 billion of foreign exchange reserves to start 2018. When we account for sovereign and PDVSA debt that will be paid from August 2017 through December, however, we subtract $3.8 billion to receive an estimated foreign exchange reserves level of $5.26 billion on January 1st.

There are many caveats to this forecasting methodology, of course, a primary flaw being that the world of finance (and the world as a whole) generally does not move linearly. If it did, we wouldn't need posts like this trying to make predictions, because everything would be so linear we could forecast decades and decades into the future. An additional threat to the integrity of this model is that Venezuela could strike further bond deals, as it did controversially with Goldman Sachs earlier this year.

However, using duct tape to fix a crumbling building can only do so much, per se. Venezuela will continue to struggle amidst this economic, political, social and broadly-humanitarian crisis. With the Venezuelan bolivar recording a 750%+ depreciation over the last year as of this typing, a currency crisis appears to be blooming. This is quite ominous, as currency crises historically have been seen to precede sovereign debt crises. Investors appear to be baking such a potential credit event into their investments, with the Venezuela five-year credit default swap spread reaching levels not seen since 2016 this past Thursday (Figure 1 below).

Figure 1: Venezuela's 5-year CDS spread.
Chart from http://www.boursorama.com/bourse/cours/graphiques/historique.phtml?symbole=3xVENZ

The vultures continue to circle around Venezuela, and while a credit event may not occur for some time, it's quite apparent that the economic & financial situation, much less the social and political situation, is unsustainable.

Andrew

Economic Data Continues to Suggest Late-Stage Business Cycle

A review of economic data continues to suggest the United States is in the late stages of the current economic expansion of the broader business cycle. We'll begin analysis of the economic data with monthly job openings for total non-farm jobs.

Source: FRED of St. Louis

Shown above are total non-farm job openings for the United States since just before the recession that began at the turn of the century. Increasing job openings indicates increasing opportunities for the unemployed to find a job, and thus signals a healthy (or at least improving) economy. This phenomenon is seen in the aftermath of both the dot-com bubble recession and the 2007-2009 recession (shaded gray areas).

I input a trend line (red), from the nadir in job openings after the 2007-2009 recession to around the value reached in April 2017. This is to show the general slope of job openings so far, and can be used practically by watching for any sudden deviations from the trend line, for example if the number of job openings were to suddenly skyrocket or plummet.
I also input a shorter-term trend line (green), which begins January 2015 and ends May 2017, the last recorded data point. Note how the slope of this new line is notably lower than the slope seen throughout the entire economic expansion. While we can visually see the values still oscillate around that red slope line, the recent trend has been to see momentum in job openings slow. The green trend line is meant to show that while the number of job openings still appear to be expanding, indicating some slack in the labor market, the momentum of this upward trend may be beginning to falter, a typical symptom of the late stage of an economic expansion of the business cycle.

One data set is not nearly enough to validate such a claim, of course, so we'll now expand our view into a few more parameters.

Source: FRED of St. Louis

To broaden the scope of parameters that generally track the business cycle, I created a graph composed of total non-farm job openings, levels of commercial and industrial (C&I) loans from all commercial banks, total vehicle sales, and the unemployment rate. These parameters are shown in the lines colored blue, red, green and purple, respectively. Additionally, note that these lines do not use the raw data numbers (e.g. job openings in thousands of people or unemployment rate in percentage), but instead are made into an index. In other words, I have set each parameter to show an index value of "100" at their respective nadirs during or immediately following the 2007-2009 recession. By using this index format, we are better able to track slowing momentum and potential turning points in these business cycle-sensitive parameters.

Commercial and Industrial Loans, All Commercial Banks
Since the total non-farm job openings parameter has been analyzed above, we now take a look at commercial and industrial loans created by all commercial banks, as aggregated by the Federal Reserve. For this parameter, the index value of "100" was set for October 2010, the lowest point in C&I loans resulting from the 2007-2009 recession.

Loan growth was originally negative following the end of the recession, as seems to generally happen following economic recessions (see a similar phenomenon occur following the early-2000s recession). Loans picked back up around 2011, and has been on the uptrend since. While I have not drawn out a trend line, you can see the relatively steady upward slope in C&I loans from ~2011 to the start of 2016. From 2016 until today, however, we note that the level of commercial and industrial loans given by commercial banks has plateaued. It is possible this stems from businesses not needing any further credit, having had the last eight years to enjoy economic growth. It is possible loan growth has slowed as businesses are no longer as confident about the future to significantly invest in long-term plans via loans. There is a wide variety of possible triggers for this plateau, but no matter the true reason(s), C&I loan growth has indeed appeared to hit a plateau.
This is similarly symptomatic of a late-stage economic expansion, as companies (again, for unknown reason(s)) begin to curb their loans. Should this be a protracted phenomenon, the lack of strong investment in long-term growth plans could hamper the current economic expansion.

Total Vehicle Sales
Vehicle sales are seen as another indicator of the business cycle, for their widespread usage by consumers but also their position as a durable good. As a durable good, vehicles will generally be purchased when consumers are upbeat about the economy and have ample funds. Consequentially, vehicle sales track the business cycle. For this parameter, the index value of "100" was assigned to February 2009, the nadir of sales amidst the recession.

Similar to C&I loans, it is not difficult to visually draw a line from February 2009 to roughly mid-2015 where the growth / slope line of vehicle sales was stable and strong. While 2015 and 2016 both saw record vehicle sales, the years also saw the emergence of a plateau as vehicle sales jumped from 16.9 million in 2014 to 17.8 million in 2015, followed by 17.9 million in 2016. Analysts have noted poor vehicle sales numbers so far in 2017, at least relative to the prior record-setting year. The apparent retreat of consumers from vehicle purchases may suggest less confidence in the economy, a preference to save money, or other reasons. Again, the reasoning may be unclear, but the data once again shows a characteristic of a late-stage economic expansion.

Unemployment Level
This parameter isn't shown here to be scrutinized for a plateau so much as a slowing in momentum. Note in the early-2000s recession and the 2007-2009 recession that the unemployment rate only began rising notably in roughly the middle of each recession. Consequentially, the unemployment rate appears to be a bit of a lagging indicator. Thus, to identify a late-stage economic expansion, we would be seeking continued downward movements (signaling lower unemployment) but at a slower clip than before. The unemployment rate index was set to "100" for October 2009.

It's not difficult to see that we have entered a lower grade of momentum for a decreasing unemployment rate, the expected signal of a late-stage economic expansion. Since this is more of a lagging indicator, analysis is somewhat less clear than C&I loans or vehicle sales, but we can still see here that the labor market is approaching full employment, which would likely signify the peak of the economic expansion.


We've now gone over a handful of broad economic indicators and seen that there are some red flags pointing towards the United States currently in the ending stages of the economic expansion. A personal favorite of mine is to look at delinquency rates for various types of loans, as delinquency rates typically increase prior to the official start of a recession, as defined by the NBER. Once again, I have set the parameters to indexes and set the index value of "100" at each parameter's respective nadir during or immediately following the 2007-2009 recession.

Source: FRED of St. Louis

In this chart, I have shown a number of delinquency rates:

• Delinquency Rate on All Loans, All Commercial Banks (dark blue, index=100 at 2010 Q1)
• Delinquency Rate on Loans Secured by Real Estate, All Commercial Banks (red, index=100 at 2010 Q1)
• Delinquency Rate on Credit Card Loans, All Commercial Banks (green, index=100 at 2009 Q1)
• Delinquency Rate on Commercial and Industrial Loans, All Commercial Banks (purple, index=100 at 2009 Q3)
• Delinquency Rate on Consumer Loans, All Commercial Banks (turquoise, index=100 at 2009 Q2)

These various delinquency rates cover a rather broad area of delinquent loans (e.g. the Delinquency Rate on All Loans), but also on more focused portions of the economy (e.g. Delinquency Rate on Credit Card Loans). I won't go into detail on each individual parameter here because they're all not too dissimilar from each other, but it's quite apparent that prior to the last two recessions, we have seen delinquency rates across the board first lose downward momentum, and then begin rising in the months before the recession's official start. 

This trend of a trough, followed by rising delinquency rates prior to the official start of a recession is seen across the board for the five delinquency rates graphed above. In recent months, we have seen mixed indicators. For instance, the Delinquency Rate on All Loans and the Delinquency Rate on Loans Secured by Real Estate are exhibiting some slowing downward momentum, but the delinquency rates on credit card loans, C&I loans, and consumer loans appear to be rising. With these mixed signals not allowing us to say that a recession is definitely on the horizon, these delinquency rates do appear to show a late-stage expansion.

There are other economic data we could observe that both support and deny the assertion that we are in a late-stage expansion of the business cycle, but with the consensus seeming to revolve more around the confirmation that we are in a late-stage economic expansion, it's worth more to see if that is indeed the case than to decide otherwise and potentially be burned by that ignorance.

In sum, I do believe we are in the late stage of this economic expansion- an accusation that has been made before to no avail, but this time supported by economic data. 

Andrew

Libor Increases At Quickest Month-over-Month Pace in Over 12 Months

The 3-month London Interbank Offered Rate, or Libor, has recorded its strongest month-over-month increase from the start of June to the start of July in at least twelve months.

Note: Dates are meant to signify the start of each month, and the first of each month may not necessarily have been a business day.

From the first business day of June to the first business day of July, the 3-month Libor rate increased by 8.266 basis points, from roughly 1.22% to about 1.30%. Previously, the strongest increase in this rate had been from August to September 2016, when the rate saw a rise of 7.657 basis points from ~0.76% to ~0.84%.

The magnitude of increases in that period and the current period may have been similar, but the consequences are different. An increase in the Libor rate to 1.3% at this magnitude of over eight basis points signifies tighter financial conditions, tightening at a respectable pace. To be sure, a Libor rate of 1.30% today is well below the 4.54% rate seen at the start of January 2006, just before financial markets began seizing up as the financial crisis commenced.

Thus, a Libor rate of 1.3% still reflects rather loose monetary conditions - however, these conditions are certainly tighter than those seen in fall 2016, going strictly by the three-month Libor rate. Indicators such as the St. Louis Fed's Financial Stress Index show incredibly accommodative monetary policies, ascertained by a broad variety of financial instruments and indicators. I don't disagree with this, hence the emphasis on how monetary conditions appear somewhat tighter only relative to last fall.

From the start of June 2017 to the start of July 2017, the three-month Libor rate increased at a magnitude not seen in over a year, indicating a quickening pace of tightening monetary conditions. However, when zooming out to broader indicators (e.g. St. Louis Financial Stress Index) and broader timeframes (e.g. 3-month Libor in January 2006), monetary conditions remain quite accommodative. For now, this is merely something to keep an eye on.

Andrew