US 10-Year T-Note Probability Model

Introduction

The 10-year Treasury note is a debt obligation issued by the United States government with a maturity of 10 years upon initial issuance. A 10-year Treasury note pays interest at a fixed rate once every six months and pays the face value to the holder at maturity. The 10-year Treasury yield is the current rate Treasury notes would pay investors if they bought them today. The U.S. government partially funds itself by issuing 10-year Treasury notes.

The U.S. Treasury issues 10-year T-notes at a face value of $1,000, and a coupon specifying a certain amount of interest to be paid every six months. The notes are sold to institutional investors, like banks and other financial companies, through auctions conducted by the Federal Reserve. Institutions then resell these notes to investors in the secondary market, where the yield is determined.

The 10-year T-note is the most widely tracked government debt instrument in finance. Its yield is often used as a benchmark for other interest rates, like those on mortgages and corporate debt, though commercial interest rates do not track the 10-year yield exactly. The yield on the 10-Year Note is the most commonly used Risk-Free Rate for calculating a company’s Weighted Average Cost of Capital (WACC) and performing Discounted Cash Flow (DCF) Analysis.

The 10-year T-note can also impact the stock market, with movements in yield creating volatility. Rising yields may signal investors are looking for higher return investments but could also spook investors who fear that the rising rates could draw capital away from the stock market. Falling yields suggest that corporate borrowing rates will also decline, making it easier for companies to borrow and expand, thus giving equities a boost.

Global events can also have a significant impact on Treasury yields—U.S. Government bonds are considered the safest investments anywhere in the world, and when geopolitical events create upheaval, Treasuries are often in high demand from international investors, leading to lower yields.

Changes in the movement of 10-year Treasury yield tell us a great deal about the economic landscape and global market sentiment. Declines in the 10-year Treasury yield indicate caution about global economic conditions while gains signal global economic confidence. Thus, demand for the 10-year US Treasury Note can show investor confidence in the state of the economy. When investors have high confidence in the performance of the economy, they look for investments with a higher return than the 10-year Treasury Note. This triggers a drop in the price of the T-Note (a rise in its yields), reflecting the lower level of demand. In contrast, when investors have low confidence in the state of the economy, the demand for safer, government-backed 10-year T-notes increases, resulting in T-Note price increase (a fall in its yields.) The prices of less secure investments will decline because of their higher risk of default.

Probability Model

Apart from the use of normal macro-economic and technical analysis that can be conducted on the 10-year T-Note, we also like to examine probabilities of directional reversal in the yield. We deploy the methodology similar to all the other stock-index probability models we run for clients, albeit at a simpler level. Since the 10-year T-note is a macroeconomic time-series, it is not as volatile as stock indices and therefore lends itself very well to this sort of analysis.

The first objective is to identify historical up and down trends to construct the probability database of declines/advances over time. For this purpose, we have found the best results using weekly Open-High-Low-Close (OHLC) candle data series for the 10-year T-Note. We also determined that the 13-period (quarterly) Zig-Zag indicator does a sterling job of picking out the major up and down trends, with no rally or correction smaller than 10%, as depicted below.

Note the chart above depicts weekly closing prices which can sometimes have different high/low points to OHLC data so there is not always perfect alignment with tops and bottoms as there would be with the OHLC data used by the probability model.

From this trend demarcation, we can create the historical database of OHLC trough-to-peak gains and peak-to-trough losses, which are shown below sorted by size and not in chronological order:

During any current uptrend or downtrend in the US 10-year T-note, the model uses the gains/losses computed from highs and lows in the OHLC data and compares it to the historical data in the above two charts to determine how many times in the past gains/losses were worse than those currently observed. This shows us which percentile gain/loss we are in and from this we can infer the probability of the current trend ending.

As an example, as of 24/04/2022 the current up-leg of the US 10-year T-note yield has made a 156.9% gain. This has only been exceeded once from the historical record observed in the first chart above, implying we have only witnessed a larger gain 1/29 =3.4% of the time. We can convert this to a percentile via 100-3.4 = 96.6% and infer there is a 96.6% probability we are near the end of the current up-trend.

With the passing of time, as up and down trends come and go, more data is added to the database, allowing our probability density to be more granular and to provide for better statistical relevance.

The results for this computation are shown below for the 38 years since 1984:

We note from the above that minor changes in trend, as in the 1988-1992 period, do not provide probabilities that are useful to determine risk or opportunity. It is the larger moves that prove more useful, those generating probabilities of 70% or more (the higher the better.) Therefore, this method will not allow you to capture all potentially profitable moves but is a useful risk measurement tool nonetheless.

Another thing to note is that in 2020, the 10-year T-Note bounced from a near-zero value to make the largest percentage gain ever for the historical period under review. This of course resulted in the probability shooting up to 100% very quickly and remaining there until the peak was made. We conclude that large moves up or down that are much higher than those in the historical database will also therefore provide probabilities that are not useful.

Finally, these probabilities are not actionable signals. You still need to use your traditional methods (trend lines, technical analysis etc.) to determine buy or sell actions and these probabilities merely provide contextual information in the form of levels of convictions of those actions.

Customer access

Subscribers to our Standard & PRO subscriptions can access the probability model from the WEEKLY CHARTS menu, in the US10Y tab as shown below:

You can see the bounce in T-Note yields in late March 2020, coming off near-zero yields that obviously made the percentage gains in yields skyrocket and hence the probabilities locking almost immediately on 90% until the next peak was locked-in.

About RecessionALERT

Dwaine has a Bachelor of Science (BSc Hons) university degree majoring in computer science, math & statistics and is a full-time trader and investor. His passion for numbers and keen research & analytic ability has helped grow RecessionALERT into a company used by hundreds of hedge funds, brokerage firms and financial advisers around the world.
Comments are closed.

  THU 29 SEP : CERTAIN PROBABILITY MODELS TO BE DISCONTINUED FROM 1 OCT