Rational Learning Research Articles

Learning How to Smile: Can Rational Learning Explain the Predictable Dynamics in the Implied Volatility Surface?

We develop a general equilibrium asset pricing model under incomplete information and rational learning to explain the yet unexplained predictability of option prices. In our model, the fundamental dividend growth rate is unknown and subject to breaks, with time periods between breaks that follow a memory-less stochastic process. Immediately after a break there is insufficient information to price option contracts accurately. However, as new information arrives a representative Bayesian agent recursively learns about the parameters of the process followed by fundamentals. We show that learning makes beliefs time-varying in ways that induce large and dynamic risk premia in option prices and their implied volatilities. In addition, we find that learning generates different effects across option contracts with different strike prices and maturities. This induces realistic movements in the volatility surface implicit in option prices.

Limited information minimal state variable learning in a medium-scale multi-country model

Rational expectations has been the dominant way to model expectations, but the literature has quickly moved to a more realistic assumption of boundedly rational learning where agents are assumed to use only a limited set of information to form their expectations. A standard assumption is that agents form expectations by using the correctly specified reduced form model of the economy, the minimal state variable solution (MSV), but they do not know the parameters. However, with medium-sized and large models the closed-form MSV solutions are difficult to attain given the large number of variables that could be included. Therefore, agents base expectations on a misspecified MSV solution. In contrast, we assume that agents know the deep parameters of their own optimising frameworks. However, they are not assumed to know the structure nor the parameterisation of the rest of the economy, nor do they know the stochastic processes generating shocks hitting the economy. In addition, agents are assumed to know that the changes (or the growth rates) of fundament variables can be modelled as stationary ARMA(p,q) processes, the exact form of which is not, however, known by agents. This approach avoids the complexities of dealing with a potential vast multitude of alternative misspecified MSVs.Using a new multi-country euro area model with boundedly estimated rationality we show that this approach is compatible with the same limited information assumption that was used in deriving and estimating the behavioural equations of different optimising agents. We find that there are strong differences in the adjustment path to the shocks to the economy when agents form expectations using our learning approach compared to expectations formed under the assumption of strong rationality. Furthermore, we find some variation in expansionary fiscal policy in periods of downturns compared to boom periods.

Almost Everybody Disagrees Almost All the Time: The Genericity of Weakly Merging Nowhere

Suppose we randomly pull two agents from a population and ask them to observe an unfolding, infinite sequence of zeros and ones. If each agent starts with a prior belief about the true sequence and updates this belief on revelation of successive observations, what is the chance that the two agents will come to agree on the likelihood that the next draw is a one? In this paper we show that there is no chance. More formally, we show that under a very unrestrictive definition of what it means to draw priors ?randomly,? the probability that two priors have any chance of weakly merging is zero. Indeed, almost surely, the two measures will be singular?one prior will think certain a set of sequences that the other thinks impossible, and vice versa. Our result is meant as a critique of the ?rational learning? literature, which seeks positive convergence results on infinite product spaces by augmenting the process of Bayesian updating with seeming regularity conditions, variously labeled ?consistency? or ?compatibility? assumptions. Our object is to investigate just how regular these assumption and results are when considered in the space of all possible prior distributions. Our results on the genericity of nowhere weak merging and singularity speak not just to the specific assumptions and results that appear in the literature, but to the ?rational learning? approach generally. We call instead for a different approach to learning, one that recognizes the necessity of genuine, substantive restrictions on beliefs and proposes ?extra rational? restrictions that are explicitly grounded in our best understanding of human behavior, ideally gleaned from experimental data.

Framing pre-service teachers’ professional learning using Web2.0 tools: positioning pre-service teachers as agents of cultural and technological change

This article addresses key issues in pre-service teachers’ professional learning. The argument explores pre-service teachers’ learning and practice, which is both informed by technology and which uses technologically enhanced practices in classrooms as learning and teaching strategies. The article is contextualized by current international debates regarding technicist and rational post-modern professional learning cultures, performativity and national standards. The article positions pre-service teachers as potential agents of change for schools.

When the Claim Hits: Bilateral Investment Treaties and Bounded Rational Learning

Using the international investment regime as its point of departure, the article applies notions of bounded rationality to the study of economic diplomacy. Through a multimethod approach, it shows that developing countries often ignored the risks of bilateral investment treaties (bits) until they themselves became subject to an investment treaty claim. Thus the behavior of developing country governments with regard to the international investment regime is consistent with that routinely observed for individuals in experiments and field studies: they tend to ignore high-impact, low-probability risks if they cannot bring specific “vivid” instances to mind.

A parsimonious model of subjective life expectancy

This paper develops a theoretical model for the formation of subjective beliefs on individual survival expectations. Data from the Health and Retirement Study (HRS) indicate that, on average, young respondents underestimate their true survival probability whereas old respondents overestimate their survival probability. Such subjective beliefs violate the rational expectations paradigm and are also not in line with the predictions of the rational Bayesian learning paradigm. We therefore introduce a model of Bayesian learning which combines rational learning with the possibility that the interpretation of new information is prone to psychological attitudes. We estimate the parameters of our theoretical model by pooling the HRS data. Despite a parsimonious parametrization we find that our model results in a remarkable fit to the average subjective beliefs expressed in the data.

A General Framework for Rational Learning in Social Networks

This paper provides a formal characterization of the process of rational learning in social networks. Agents receive initial private information and select an option out of a choice set under uncertainty in each of infinitely many periods, observing the history of choices of their neighbors. Choices are made based on a common behavioral rule. Conditions under which rational learning leads to global consensus, local indifference and local disagreement are characterized. In the general setting considered rational learning can lead to pairs of neighbors selecting different actions once learning ends while not being indifferent among their choices. The effect of the network structure on the degree of information aggregation and speed of convergence is also considered and an answer to the question of optimal information aggregation in institutions provided. The results highlight distinguishing features between properties of Bayesian and non-Bayesian learning in social networks

A general framework for rational learning in social networks

This paper provides a formal characterization of the process of rational learning in social networks. Agents receive initial private information and select an action out of a choice set under uncertainty in each of infinitely many periods, observing the history of choices of their neighbors. Choices are made based on a common behavioral rule. Conditions under which rational learning leads to global consensus, local indifference, and local disagreement are characterized. In the general setting considered, rational learning can lead to pairs of neighbors selecting different actions once learning ends while not being indifferent among their choices. The effect of the network structure on the degree of information aggregation and speed of convergence is also considered, and an answer to the question of optimal information aggregation in networks is provided. The results highlight distinguishing features between properties of Bayesian and non-Bayesian learning in social networks.

Rational learning strategy behind irrational behavior(The 30th Annual Meeting)

Rational learning strategy behind irrational behavior(The 30th Annual Meeting)

Planning for learning and learning about planning in social work fieldwork

Fieldwork education is a crucial component of social work education. Many social work students regard their placement experiences as the most profound learning experiences of their studies. The students undertake their field placements in a diverse range of organisational contexts, and in so doing perform a myriad of tasks, adopt a variety of roles, implement a range of practices, and engage with numerous people. Needless to say, social work students have a rich set of learning opportunities within such diversity. An important part of the fieldwork process is the development of learning plans; these plans guide and direct the students’ roles, tasks and learning, and are often an important framework by which assessment of competency and learning takes place. However, learning plans presuppose a logical and conceptual clarity, which needs to be learned if they are to be functional and effective documents. This then poses many challenges in relation to how students might develop a learning plan for fieldwork. This paper explores some of the problems, and offers practical guidance, for students and fieldwork educators to develop rational learning plans in diverse and complex contexts.

Soviet communism and the British planning movement: rational learning or Utopian imagining?

This article examines the impact of Soviet communism on the British planning movement, especially from the 1930s–1960s. It is a very specific case study of international knowledge flows. The broad question of whether these resulted from a consciously rational learning process or were a more culturally constructed exercise in ‘imaginative geography’ informs the empirical investigation. This focuses on the evolving political context for Anglo-Soviet planning contacts, the shifting motivations for seeking Soviet knowledge and the evolving balance of resultant relationships. The important role of a ‘friendship organisation’, the Society for Cultural Relations with the USSR, is examined. From 1955, the British Council created a new basis for contact that was less sympathetic to Soviet ideals. Many specific instances of planning connections are identified, especially visits in the 1930s and during the ‘Khrushchev thaw’ in the 1950s. Overall, predominantly Utopian 1930s' views of Soviet urban planning changed towards more critical and negative views. Throughout, however, even the most sympathetic architects and planners retained some critical detachment and never proposed direct copying of Soviet planning models. The main impact for British planning was as a vast case study of the virtues of a bold, comprehensive and state-led approach.

Boundedly rational banks’ contribution to the credit cycle

We investigate how banks’ boundedly rational learning influences their views about default risks over the business cycle. Our analysis details the direction and the magnitude of these effects assuming that banks update probability in a Bayesian way. With a limited experience span lenders are liable to overestimate (underestimate) losses from defaulting loans early (late) in the boom. Depending on their experience span, banks turn over-optimistic and underprice default risk 3–5 years into the boom. During recessions an overpricing of risk begins just quarters into the recession. Our simulations are calibrated with U.S. data and provide evidence for the view that banks contribute to excessive lending during the upswing and to credit crunches in recessions.

Effect of Regret

Regret is a proposed explanation for many puzzles in economics and finance. Yet very few studies analyze the effect of experienced regret on subsequent decisions in a real-world-setting. We find that after experiencing regret, individuals are more likely to change their decision to place a market or limit order. Confirming the predictions of regret theory, the effect of regret is stronger following an action rather than inaction, loss on the prior order, and an unusual order strategy for the individual. Moreover, decisions influenced by regret yield poor returns. The totality of these results rules out rational learning as an explanation.

Pay for Performance from Future Fund Flows: The Case of Private Equity

Lifetime incomes of private equity general partners (GPs) are affected by their current funds’ performance not only directly, through carried interest profit-sharing provisions, but also indirectly by the effect of the current fund’s performance on GPs’ abilities to raise capital for future funds. In the context of a rational learning model, which we show better matches the empirical relations between future fund-raising and current performance than behavioral alternatives, we estimate that indirect pay for performance from future fund-raising is of the same order of magnitude as direct pay for performance from carried interest. Consistent with the learning framework, indirect pay for performance is stronger when managerial abilities are more scalable and weaker when current performance is less informative about ability. Specifically, it is stronger for buyout funds than for venture capital funds, and declines in the sequence of a partnership’s funds. Total pay for performance in private equity is both considerably larger and much more heterogeneous than implied by the carried interest alone. Our framework can be adapted to estimate indirect pay for performance in other asset management settings. (JEL G11, G23, J33)

Learning to Represent a Multi-Context Environment: More than Detecting Changes

Learning an accurate representation of the environment is a difficult task for both animals and humans, because the causal structures of the environment are unobservable and must be inferred from the observable input. In this article, we argue that this difficulty is further increased by the multi-context nature of realistic learning environments. When the environment undergoes a change in context without explicit cueing, the learner must detect the change and employ a new causal model to predict upcoming observations correctly. We discuss the problems and strategies that a rational learner might adopt and existing findings that support such strategies. We advocate hierarchical models as an optimal structure for retaining causal models learned in past contexts, thereby avoiding relearning familiar contexts in the future.

When the Claim Hits: Bilateral Investment Treaties and Bounded Rational Learning

Using the international investment regime as its point of departure, the paper introduces notions of bounded rationality to the study of economic diplomacy. Through a multi-method approach, it shows that developing countries often ignored the risks of bilateral investment treaties (BITs) until they themselves became subject to an investment treaty claim. Thus the behavior of developing country governments with regard to the international investment regime is consistent with that observed for individuals in experiments and field studies: they tend to ignore high-impact, low-probability risks if they cannot bring specific ‘vivid’ instances to mind.

Representations of Old Age in Falls Prevention Websites: Implications for Likely Uptake of Advice by Older People

Introduction: The representations of old age in falls prevention websites were explored and considered for their potential impact on older people's uptake of the advice. Method: Websites were searched for using the strategy of a previous systematic-style review and evaluated using the principles of discourse analysis. Findings: In the analysis of 33 websites, three main subject positions afforded to older readers were identified: (1) Passive recipients: victims of the ageing process, ignorant, ill-informed and vulnerable; (2) Rational learners: responsive to information, rational problem solvers and compliant with prescriptive advice; and (3) Empowered decision makers: actively engaged with and evaluative of information, autonomous and responsible for their course of action. Discussion: Falls prevention websites were rarely designed according to evidence-based recommendations concerning fit with positive self-identity and empowerment of active self-management of health. Although the representation of older people as passive and inert was most evident, the image of empowered decision makers was most likely to engage older people in preventing falls. Conclusion: Occupational therapists should ensure that they represent older people in a positive and respectful manner in falls prevention information, both that available through the internet and in written form.

Research on a Multi-Objective Genetic Algorithm for Rational Agent Learning Model

As a software component which is capable of learning in an autonomous way, software agent should have the capability of learning in a dynamic environment. Genetic Algorithm has a wide perspective in the machine learning because of its unique characteristic (e.g. dynamic adaptability, self-organization, global convergence and robustness). But when applying GA to agent’s dynamic learning model, it encounters a series of problem. In this paper, a Modifided Multi-Objective Genetic Algorithm(MMOGA) will be introduced to solve these problems. Finally, an Agent’s Dynamic learning model based on a MMOGA which has the flexible dynamic learning capability, better global convergence and performance, will be introduced.

Learning Self-Control *

This article examines how a decision maker who is only partially aware of his temptations learns about them over time. In facing temptations, individuals use their experience to forecast future self-control problems and choose the appropriate level of commitment. I demonstrate that rational learning can be perpetually partial and need not result in full sophistication. The main result of this article characterizes necessary and sufficient conditions for learning to converge to full sophistication. I apply this result to a consumption-savings environment in which a decision maker is tempted by present bias and establish a learning-theoretic justification for assuming sophistication in this setting.“An individual who finds himself continuously repudiating his past plans may learn to distrust his future behavior, and may do something about it.” — Strotz (1955)

When global structure “Explains Away” local grammar: A Bayesian account of rule-induction in tone sequences

While many constraints on learning must be relatively experience-independent, past experience provides a rich source of guidance for subsequent learning. Discovering structure in some domain can inform a learner’s future hypotheses about that domain. If a general property accounts for particular sub-patterns, a rational learner should not stipulate separate explanations for each detail without additional evidence, as the general structure has “explained away” the original evidence. In a grammar-learning experiment using tone sequences, manipulating learners’ prior exposure to a tone environment affects their sensitivity to the grammar-defining feature, in this case consecutive repeated tones. Grammar-learning performance is worse if context melodies are “smooth” — when small intervals occur more than large ones — as Smoothness is a general property accounting for a high rate of repetition. We present an idealized Bayesian model as a “best case” benchmark for learning repetition grammars. When context melodies are Smooth, the model places greater weight on the small-interval constraint, and does not learn the repetition rule as well as when context melodies are not Smooth, paralleling the human learners. These findings support an account of abstract grammar-induction in which learners rationally assess the statistical evidence for underlying structure based on a generative model of the environment.