Liquidity and Trading in Bond Markets: Why Bonds Are So Hard to Trade

Book: Systematic Fixed Income: An Investor’s Guide Author: Scott A. Richardson, Ph.D. Publisher: John Wiley & Sons, 2022 ISBN: 9781119900139

You can have the best investment signals in the world. But if you cannot actually trade the bonds you want to buy, none of that matters. Chapter 9 of Richardson’s book is about liquidity and trading, and it is probably the most practical chapter in the entire book. The reality of bond market trading is very different from stock market trading.

Bonds vs. Stocks: The Liquidity Gap

Richardson presents a comparison that makes the problem immediately clear. US government bonds, corporate bonds, and stocks all have market capitalizations over $10 trillion USD. But average daily trading volumes tell a completely different story.

US Treasury bonds trade about $600 billion per day. US stocks trade about $500 billion per day. US corporate bonds? About $40 billion per day. That is a huge gap.

Bid-ask spreads are also revealing. Treasury bonds: less than 0.05 percent. Stocks: less than 0.1 percent. Corporate bonds: 0.3 to 0.5 percent.

But the most telling number is bid-ask spread divided by volatility. For stocks it is about 0.5 percent, for Treasuries about 1 percent, and for corporate bonds about 6 percent. Corporate bonds are by far the most expensive asset class to trade relative to how much they move. That single ratio explains a lot about why systematic fixed income investing is different from systematic equity investing.

How Bonds Actually Trade

Corporate bonds trade over-the-counter, meaning there is no central exchange. Instead, dealers act as intermediaries between buyers and sellers. Traditionally, dealers held large inventories and acted in a “principal” capacity. They would buy bonds from sellers, hold them on their books, and sell to buyers later.

But that model has changed dramatically. Primary dealers held nearly $265 billion in IG corporate bonds at the end of 2007. By 2020 it was just $4 billion. Regulatory capital requirements and electronification gutted dealer inventory.

So agency trading has become more common. The dealer acts as a matchmaker, connecting buyers and sellers. This is actually good for systematic investors because they have views on thousands of bonds and can respond quickly.

IG vs. HY Trading Conventions

Investment-grade bonds are quoted on a spread basis. A dealer might offer to buy at 83 basis points over Treasuries and sell at 78. Because of the inverse relationship between yields and prices, a higher spread means a lower price. So the dealer’s “bid” is the higher spread number. High-yield bonds are quoted on a price basis, more like stocks.

The difference in liquidity between individual bonds is enormous. A liquid IG bond (like GE) has multiple dealers quoting tight spreads under 5 basis points with substantial size. A less liquid IG bond might have fewer quotes, wider spreads, smaller sizes, and stale quotes.

The top 50 stocks in the S&P 500 trade more than 50,000 times per day. The top 50 corporate issuers might trade around 20 times per day, but at much larger sizes (about $2 million per trade versus under $20,000 for stocks).

Credit Index Instruments

For market-level views on credit, investors have three main alternatives to individual bonds.

Credit default swap indices (CDX for North America, iTraxx for Europe) are equally weighted default baskets of corporate issuers. They trade on swap execution facilities with tight spreads and fast execution. Most liquidity concentrates in the five-year “on-the-run” contract. The limitation is tracking error relative to cash bond indices.

ETFs like JNK (high yield) or LQD (investment grade) trade like stocks with streaming prices. But they have management fees (up to 0.5 percent annually) and tend to lag the underlying indices. There is also a create-and-redeem mechanism where baskets of bonds can be converted to ETF units and vice versa.

Total return swaps have low tracking error but come with high costs and counterparty risk.

Electronic Trading Is Growing

Corporate bond trading was almost entirely voice-based for a long time. As of December 2020, about 38 percent of US IG and 27 percent of US HY volumes were on electronic platforms (compared to 65 percent for interest rate markets).

The main platforms are Bloomberg (dominant in Europe) and MarketAxess (dominant in North America, about 80 percent share), plus TradeWeb, TruMid, and LiquidNet.

Trading protocols include anonymous requests for quote (51 percent of IG volumes), disclosed RFQs (20 percent), auctions, central limit order books, portfolio trading, and streaming. We are still far from a stock market-style exchange, but the trend is clear.

Single-name CDS contracts have evolved faster because they are standardized and centrally cleared. Systematic strategies already account for about 25 percent of CDS trading volumes.

Primary Markets: The New Issue Concession

When corporations issue new bonds, there is a “new issue concession.” Newly issued IG bonds earn about 0.25 percent in the first couple of days, and HY bonds earn about 0.85 percent. Returns are positive for 72 percent of IG new issues and 86 percent of HY new issues.

The primary market is heavily intermediated. The top five dealers account for about 49 percent of US IG underwriting. The timeline from pricing talk to allocation is compressed into a single day.

Systematic investors are well positioned here. You already have views on thousands of issuers, so you can quickly evaluate new issues. And you can build credibility by communicating that your investment horizon is measured in months, not days.

Liquidity Provision and Axes

Smart systematic investors do not just take liquidity. They provide it. Dealers communicate their interest in specific bonds through “axes,” broadcast widely or shared privately. The axe typically shows pricing inside the prevailing bid-ask spread.

A systematic process is uniquely positioned here. You already have a view on every bond, updated frequently. When a dealer shows you an axe, you can quickly decide if it improves your portfolio. This lowers transaction costs and allows continuous rebalancing.

Transaction Cost Analysis

Measuring your trading costs properly is critical. The total “shortfall” has two components: actual trading costs (difference between executed price and the pre-trade mid-price) and slippage (the price drift between when your model generated the trade list and when you actually traded).

Slippage can come from many sources: different pricing data sources, time spent on data checking, and the delay while waiting for liquidity. After the trade, “markout” analysis tracks how prices move to detect adverse selection.

Richardson makes the practical point that you cannot go back in time and timestamp your order management steps. Committing to proper data storage early is a prerequisite for good transaction cost analysis later.

Bond Substitutions

Sometimes you simply cannot trade the bond you want. In that case, you need a systematic substitution process: identify the next-best bond, adjust the trade size for different risk contribution, and execute. Traders provide live liquidity information and trigger substitution rules when predefined thresholds are hit. Having views on thousands of bonds means you always have a ready alternative.

This post is part of a series retelling Systematic Fixed Income: An Investor’s Guide by Scott A. Richardson.

Previous: Portfolio Construction in Fixed Income Next: Sustainable Fixed Income Investing