{"id":45180,"date":"2023-04-24T21:11:50","date_gmt":"2023-04-25T02:11:50","guid":{"rendered":"https:\/\/prefblog.com\/?p=45180"},"modified":"2023-04-24T21:11:50","modified_gmt":"2023-04-25T02:11:50","slug":"april-24-2023","status":"publish","type":"post","link":"https:\/\/prefblog.com\/?p=45180","title":{"rendered":"April 24, 2023"},"content":{"rendered":"
\"\"<\/a><\/div>\n

TXPR closed at 551.68<\/a>, down 0.92% on the day. Volume today was 1.31-million, above the median of the past 21 trading days.<\/p>\n

CPD closed at 10.97<\/a>, down 0.63% on the day. Volume was 43,660, far below the median of the past 21 trading days.<\/p>\n

ZPR closed at 9.10<\/a>, down 0.76% on the day. Volume was 91,610, third-lowest of the past 21 trading days.<\/p>\n

Five-year Canada yields down to 3.09%<\/a> today.<\/p>\n

The BoC has released<\/a> a Staff Working Paper by John Duffy, Janet Hua Jiang and Huan Xie titled Pricing Indefinitely Lived Assets: Experimental Evidence<\/a>:<\/p>\n

We study indefinitely lived assets in experimental markets and find that the traded prices of these assets are, on average, about 40% of the risk-neutral fundamental value. Neither uncertainty about the value of total dividend payments nor horizon uncertainty about the duration of trade can account for this low traded price. An Epstein and Zin (1989) recursive preference specification that models the dynamic realization of dividend payments and incorporates risk preferences can rationalize the low traded price observed in our indefinitely lived asset market.
\n…<\/b>
\nIn this paper, we study the trade of assets in an experimental market with indefinite horizons, consisting of an unknown number of periods. The first period begins with trade in the asset. Following trade, each unit of the asset pays its holder a fixed dividend. Thereafter, with a constant probability \u03b4, traders\u2019 holdings of the asset carry over to the next period, and in each new period, trade in the asset takes place and asset holders earn dividends per unit held. With probability 1 \u2212 \u03b4, the asset ceases to exist; the asset market shuts down and the asset has no continuation value. This indefinite-horizon, or random-termination, design, initially proposed by Roth and Murnighan (1978), is the most commonly used approach in the laboratory to implementing infinite horizons with discounting.<\/p>\n

Unlike most finite-horizon asset markets where the FV of the asset decreases over time, the stationarity associated with indefinite horizons implies that the FV of the indefinitely lived asset is constant over time.1 The stationarity associated with indefinite horizons may be a more natural setting for understanding asset pricing decisions.2<\/p>\n

In our baseline treatment (treatment A), subjects trade in indefinite-horizon asset markets implemented by random termination (more precisely, a modified version of the block random termination scheme of Fr\u00e9chette and Yuksel (2017)). In each period the market is open, subjects first trade units of a single asset. Once trading is concluded,
\nthey receive dividend payments for each asset share they hold. Finally, a random number determines whether the asset market will continue to a new period. In each session, subjects participate in three indefinite-horizon markets (with different pre-drawn market lengths) to reveal the effect of experience, as in Smith et al. (1988). We find that traded prices are quite low, averaging around 40% of the standard FV, and they remain low even as traders gain experience. This result is rather surprising given that the vast majority of experimental asset market studies following the Smith et al. (1988) design find asset price bubbles, or prices greatly in excess of the standard FV, in the first market played, with approximate convergence to the standard FV within three market repetitions.
\n…<\/b>
\nAs a result, we conclude that neither uncertainty about the trading horizon nor uncertainty regarding total dividend payoffs can account for the low traded prices observed in the baseline treatment A relative to the other two treatments. Instead, the experimental results suggest that the dynamic realization of dividend payments plays a critical role in accounting for the low traded price in treatment A relative to the other two treatments. In treatment A, in each trading period, subjects receive dividend payments in the current period and face an uncertain continuation value in the future. In the other two treatments, as all dividend realizations are realized after the trading is completed, subjects are more likely to view the asset as a static lottery and care about the total dividend payments.<\/p><\/blockquote>\n

I’ve kind of butchered this here, not including a description of how “Treatment B” and “Treatment C” differ from “Treatment A”, so those interested will just have to read the paper. Still, this does offer food for thought in the context of the current low trading prices of the current FixedReset market; and of the absurdly high prices during the issuance boom of the early 2010s.<\/p>\n

Who wants to tweak this game to be like the FixedReset market? Say, after each round there are four mutually exclusive possibilities:<\/p>\n