[This blog post is an excerpt from a TSI commentary published on 26th January 2020]

Given the extent to which the US stock market is stretched in both momentum and sentiment terms, there doesn’t have to be a news-related catalyst for a significant correction. However, the mainstream financial press tries to link every move in the stock market to the current news. If a correction began last week or gets underway this week it’s likely that many fingers of blame will point to the Wuhan virus.

The situation is ‘fluid’, but at last count 13 Chinese cities had been placed under full or partial lockdown in an effort to prevent the virus from spreading. Furthermore, the number of countries with confirmed cases of the virus is growing.

In terms of global economic impact we think that the Wuhan virus will prove to be a minor issue, but if the number of confirmed cases continues to rise then many market participants could sell first and ask questions later. Based on what happened with similar viruses in the past, the number of confirmed cases might not peak until March.

Blame for a correction also could be directed towards the Fed, largely due to a misunderstanding of the Fed’s “repo market” operations.

The net amount of ‘liquidity’ provided by the Fed to the repo market has declined over the past couple of weeks, but not because the Fed has stopped supporting this market. The money provided to the repo market is a very short-term loan that often matures within one day, so the amount of repo money provided by the Fed will reduce over time unless the Fed adds new money (makes new loans) at a rapid pace. For example, the Fed ‘pumped’ (loaned) $74 billion into the repo market last Thursday, but there was a net decline of $10 billion on the day due to the expiration of $84 billion of previous loans.

Also worth mentioning is that the amount of money provided by the Fed to the repo market cannot exceed the demand for short-term loans in this market.

We assume that the Fed intends to withdraw from the repo market over the coming months, with the very short-term money it provides via this market steadily being replaced by the semi-permanent money it adds via the asset monetisation program — the program that we aren’t supposed to call QE, even though it is mechanically identical to QE — introduced last October. If this happens it will result in a large decline in the “Repurchase Agreements” line on the Fed’s balance sheet and could result in a small decline in the total size of the Fed’s balance sheet, but it won’t be a sign that the Fed is tightening or even becoming less easy.

To further explain, note that even though many commentators lump the Fed’s repo market support program together with the Fed’s asset monetisation program to arrive at a total amount of new Fed-generated monetary inflation, the two programs are very different and should not be combined. The reason is that despite “repo” being short for repurchase, repo operations do not involve asset purchases per se. When the Fed does a “repo” it lends money (that it creates out of nothing) to a borrower and receives collateral, usually in the form of a Treasury security or a Mortgage-Backed Security (MBS), to secure the loan. When the loan is repaid, which happens one or fourteen days later depending on whether it’s an overnight loan or a term loan, the money is returned to the Fed (and is immediately extinguished) and the collateral is returned to the borrower.

Any monetary inflation caused by the Fed’s repo operations is therefore self-extinguishing within a very short time (1-14 days). However, monetary inflation caused by the Fed’s asset monetisation program could be permanent. The reason is that although the debt securities (T-Bills) purchased by the Fed with new money will mature within a few months, the Fed has said that it will re-invest (rather than extinguish) the proceeds received when the securities mature.

The upshot is that even if the total size of the Fed’s balance sheet reduces as the repo support program winds down, as long as the Fed is adding ‘permanent’ money via its asset monetisation program it is acting in an inflationary manner. Consequently, it will not be reasonable to blame a near-term stock market downturn on the Fed becoming tighter or less easy.

Just to be clear, the Fed is at least partly responsible for the fact that the stock market rose in a virtual straight line from mid-October through to 23rd January. However, the Fed is not responsible for the market pulling back from an overbought/overbullish extreme.

The crux of the matter is that regardless of the fundamentals, large and liquid markets don’t go up or down in straight lines for long. There are always corrections. Surely we don’t have to rack our brains in an effort to come up with a fundamental reason for a correction when some short-term sentiment and momentum indicators are stretched to historic extremes. It’s more of a challenge to explain why a correction didn’t happen sooner.

Do gold mining stocks, as represented by the Barrons Gold Mining Index (BGMI) or the XAU or the HUI, lead gold bullion at significant turning points? According to many analysts the answer is a resounding “yes”. However, according to the historical record the answer is “sometimes, but not consistently”. I’ll go with the historical record.

There are plenty of examples of gold mining stocks leading the bullion price at a turning point from down to up or up to down, but there are at least as many examples of gold mining stocks lagging the bullion at a turning point. The most blatant example of the latter occurred in 1980. It is well known that gold bullion reached a long-term price top in January-1980, but it is less well known that gold mining stocks, as represented on the following weekly chart by the BGMI, experienced a huge rally after the major high in the gold price and didn’t reach a long-term peak of their own until September-1980.

Another example of gold bullion leading the gold-mining sector at a meaningful turning point occurred during the mid-1980s. First, gold bullion reached a multi-year bottom in February-1985, but it wasn’t until July-1986 that the XAU reached a similar bottom. Second, the XAU commenced an intermediate-term rally in late-July of 1986, but as noted on the following daily chart the associated bottom in the gold price occurred about 6 weeks earlier.

Here are examples of the same lead-lag relationship from the more recent past:

1) Gold bottomed on a long-term basis in early-December of 2015, but the equivalent low for the HUI didn’t occur until mid-January of 2016. Gold then reached an intermediate-term top in July-2016, but the associated top for the HUI wasn’t put in place until about a month later. In other words, the gold-mining sector lagged the bullion market at both the start and the end of the 2016 intermediate-term rally.

2) Gold made an intermediate-term bottom in August-2018, whereas the HUI waited until September-2018 to make an intermediate-term bottom of its own. Then, in May of 2019 the HUI made a sequence of lower correction lows while gold bullion remained above the low it made at the beginning of the month.

In my experience, rather than labour under the unreliable assumption that gold stocks lead the bullion at turning points, any divergence or non-confirmation between gold stocks and gold bullion should be viewed as potentially bullish after prices have become stretched to the downside and potentially bearish after prices have become stretched to the upside. For example, after prices have been trending downward for several months, a new multi-month low in the gold price that isn’t quickly confirmed by the HUI or a new multi-month low in the HUI that isn’t quickly confirmed by the gold price should be viewed as a bullish sign. By the same token, after prices have been trending upward for several months, a new multi-month high in the gold price that isn’t quickly confirmed by the HUI or a new multi-month high in the HUI that isn’t quickly confirmed by the gold price should be viewed as a bearish sign.

To put it more simply, non-confirmations between the gold price and the related mining indices matter, but it doesn’t matter whether the non-confirmations involve relative strength or relative weakness in the mining sector.

[This post is an excerpt from a recent TSI commentary]

Over the past four months the Fed has added about $400B to its balance sheet. To put this into perspective, since early September the Fed has expanded its balance sheet at an annualised rate of around 30%. According to the Fed, the purpose of this dramatic monetary expansion was to address a temporary liquidity issue in the “repo” market. The question is: If the Fed is dealing with only a temporary shortage of liquidity in the market for short-term money, why did it introduce a program in mid-October to supplement the temporary injections of “repo” money with $60B/month of permanent money?

The answer is that the Fed is dealing with something more than a temporary shortage of liquidity in the market for short-term money. The fact that the Fed sees the need to remove $60B/month of Treasury supply from the market in addition to the Treasury supply that is being removed on a temporary basis via “repo” operations implies that the overall demand for Treasury debt is falling short of Treasury supply at the Fed’s targeted interest rates. Looking from a different angle, it is clear that at current interest rates the global financial system wants more dollars and less Treasury debt. The Fed is accommodating this desire by increasing the supply of dollars to the market and reducing the supply of government debt that must be absorbed by the market.

The key phrase in the above paragraph is “at current interest rates”. If the supply of and the demand for money and credit were permitted to balance naturally then interest rates would now be much higher. However, the Fed doesn’t want supply and demand to strike a natural balance; the Fed has decided that it wants the price of credit at a certain level and that it will use its power to create and destroy money to override natural market forces. In this regard the current situation is unusual only in degree, because the Fed has been attempting to override market forces for more than 100 years.

The US Federal government is not about to slow the pace at which it emits new debt. On the contrary, the rate of growth in government debt supply looks set to rise. Therefore, one of two things will have to happen if interest rates are to stay near current low levels: The Fed will have to keep absorbing Treasury supply at a rapid pace or the market’s desire to hold Treasury debt will have to increase substantially. The latter could occur in response to a sizable decline in the US stock market or a crisis outside the US.

Within a week of its mid-October announcement we wrote that the Fed’s promise to inject $60B/month of new ‘permanent’ money was a potential game-changer, in that it could extend the current cycle (prolong the equity bull market) and lead to more “price inflation” than earlier programs. We continue to think that a cycle extension could be on the cards, but if so the recession warnings that were generated by leading indicators during the second half of last year must disappear within the next couple of months.