Foresight Bias Research Articles

Lower bounds for American option prices with control variates

This article discusses the application of optimally sampled control variates in the context of the Least-Squares Monte Carlo algorithm for pricing American options. We demonstrate theoretically that optimal sampling introduces bias when estimated exercise times are not stopping times. Numerical results show that this bias is an accurate proxy for the positive foresight bias of American option price estimates, effectively yielding in-sample lower-bound estimates.

Single Simulation Lower Bounds for Bermudan Derivatives

When pricing Bermudan derivatives by regression-based methods, foresight bias will appear in lower bounds when using a single simulation to estimate the exercise strategy and to compute lower bounds. In this paper, we propose a new method to remove this kind of bias without introducing an independent simulation. Numerical results indicate that the goodness of our method is comparable to that of using independent simulations. In addition, this method can be parallelized and enhanced by local regression. These improvements boost the accuracy and the time efficiency of lower bounds.

The linguistic and cognitive factors associated with lexical-gustatory synesthesia: A case study

Past research on lexical-gustatory synesthesia (people who associate words with tastes) has shown that linguistic factors underlie the inducer (word)–concurrent (taste) mappings in these individuals. The developmental cognitive model envisioned by Simner and Haywood (2009), and an extension of it proposed in this paper can be used to understand these linguistic associations. The first aim of the current case study was to test these models by examining the linguistic factors associated with this type of synesthesia. The influence of lexical-gustatory synesthesia on memory was investigated using a paired-associate learning task. Further, the present case study also examined if word-pairs with same or similar tastes and word-pairs with dissimilar tastes enhanced learning relative to word pairs that had no tastes. As predicted, the findings revealed possible phonological, phonological–lexical, and lexical–semantic factors linking the inducer–concurrent pairings. These findings are in line with the developmental cognitive models of LG synesthesia. There were no effects of synesthesia on memory as demonstrated by the lack of any significant difference between the synesthete and non-synesthete controls on a paired-associate learning task. Moreover, no significant differences emerged between the “no taste” and “taste” conditions (although she performed slightly better on the “no taste” condition). Interestingly, a metamemory task (judgment-of-learning) revealed the opposite. That is, the synesthete predicted that her learning would be better in the “taste” condition when compared to the “no taste” condition. This indicates that the word-pairs, which produced tastes, could have created a “foresight bias”. This is attributed to the unidirectional nature of this individual’s LG synesthesia. This finding should, however, be treated with caution because it is a preliminary finding based on a single subject and needs to be corroborated with future studies on other lexical-gustatory synesthetes.

Testing facilitates the regulation of subsequent study time

We examined how testing potentiates self-regulated learning and alleviates the foresight bias—an illusion of competence that arises from information being present during study but absent at test—and whether such benefits can transfer to non-tested material. After studying paired associates that varied in difficulty, participants either restudied or were tested on all the pairs (Experiment 1); were tested on only half of the pairs (Experiment 2); or were tested on half of the pairs and restudied the remaining pairs (Experiments 3 and 4). All items were then restudied at participants’ own pace before a final cued-recall test. In Experiment 1, interim tests enhanced the effectiveness of subsequent study time and alleviated the foresight bias, whereas interim restudying had no such benefits. Experiments 2, 3, and 4 demonstrated that such test-potentiated self-regulated learning can transfer to non-tested items if restudied intermixed with items that were tested. The results demonstrate yet another practical benefit of testing and suggest that retrieval practice can foster metacognitive sophistication among learners, serving as an experiencebased debiasing procedure.

Illusions of competence for phonetically, orthographically, and semantically similar word pairs.

Illusions of competence are thought to arise when judgements of learning (JOLs) made in the presence of intact cue-target pairs during study create a "foresight bias," such that JOLs are inflated by the apparent association between a cue and a target, despite the lack of benefit this association has for recall performance. For example, Castel, McCabe, and Roediger (2007) recently demonstrated an illusion of competence for identical word pairs (mouse-mouse). In two experiments, the authors examined possible sources for this over confidence, including phonetic, semantic, and orthographic similarity. An illusion of competence was found for homophones, synonyms, orthographically similar, and unrelated items, whereas no illusion of competence was found for word pairs with a relatively high forward-semantic association. Self-paced study times indicated that encoding fluency was not closely associated with the magnitude of over confidence. Error data revealed participants may have been engaging in strategic responding in order to maximise correct recall. Our results underscore the importance of considering factors that influence both JOLs and recall performance when considering sources of (mis)calibration in absolute accuracy.

Illusions of competence and overestimation of associative memory for identical items: Evidence from judgments of learning

The relation between subjects' predicted and actual memory performance is a central issue in the domain of metacognition. In the present study, we examined the influence of item similarity and associative strength on judgments of learning (JOLs) in a cued recall task. We hypothesized that encoding fluency would cause a foresight bias, so that subjects would overestimate recall of identical pairs (scale-scale), as compared with strong associates (weight-scale) or unrelated pairs (mask-scale). In Experiment 1, JOLs for identical word pairs were higher than those for related and unrelated pairs, but later recall of identical pairs was lower than recall of related pairs. In Experiment 2, the effect of encoding fluency (inferred from self-paced study time) was examined, and a similar pattern of results was obtained, with subjects spending the least amount of time studying identical pairs. We conclude that overconfidence for identical pairs reflects an assessment of item similarity when JOLs are made, despite associative strength being a better predictor of later retrieval.

Compressed Foresight and Narrative Bias: Pitfalls in Assessing High Technology Futures

Compressed Foresight and Narrative Bias: Pitfalls in Assessing High Technology Futures

Illusions of competence during study can be remedied by manipulations that enhance learners’ sensitivity to retrieval conditions at test

Monitoring one's knowledge during study is susceptible to a foresight bias (Koriat and Bjork, 2005). Judgments of learning (JOLs) are inflated whenever information that is present at study and absent, but solicited, at test, such as the targets in cue-target paired associates, highlights aspects of cues that are less apparent when those cues are presented alone. The present findings demonstrate that foresight bias can be alleviated by study-test experience (Experiment 1), particularly test experience (Experiments 2 and 3), and by delaying JOLs after study (Experiment 4) and that both foresight bias and its alleviation have behavioral consequences, as measured by study time allocation (Experiment 5). Collectively, the findings suggest that overconfidence and misallocation of study time arise from a mismatch that is inherent to education--that the answer is present at study and absent at test--and that alleviating the problem requires creating conditions at study that sensitize learners to retrieval conditions at test.

Mending metacognitive illusions: A comparison of mnemonic-based and theory-based procedures.

Previous research indicated that learners experience an illusion of competence during learning (termed foresight bias) because judgments of learning (JOLs) are made in the presence of information that will be absent at test. The authors examined the following 2 procedures for alleviating foresight bias: enhancing learners' sensitivity to mnemonic cues pertaining to ease of retrieval and inducing learners to resort to theory-based judgments as a basis for JOLs. Both procedures proved effective in mending metacognitive illusions-as reflected in JOLs and self-regulation of study time-but only theory-based debiasing yielded transfer to new items. The results support the notion that improved metacognition is 1 key to optimizing transfer but also that educating subjective experience does not guarantee generalization to new situations.

Exploring a mnemonic debiasing account of the underconfidence-with-practice effect.

Judgments of learning (JOLs) underestimate the increase in recall that occurs with repeated study (the underconfidence-with-practice effect; UWP). The authors explore an account in terms of a foresight bias in which JOLs are inflated when the to-be-recalled target highlights aspects of the cue that are not transparent when the cue appears alone and the tendency of practice to alleviate bias by providing learners with cues pertinent to recall. In 3 experiments the UWP effect was strongest for items that induce a foresight bias, but delaying JOLs reduced the debiasing effects of practice, thereby moderating the UWP effect. This occurred when delayed JOLs were prompted by the cue alone (like during testing), not when prompted by the cue-target pair (like during study).

Foresight Bias and Suboptimality Correction in Monte-Carlo Pricing of Options with Early Exercise: Classification, Calculation and Removal

In this paper we investigate the so called foresight bias that may appear in the Monte-Carlo pricing of Bermudan and compound options if the exercise criteria is calculated by the same Monte-Carlo simulation as the exercise values. The standard approach to remove the foresight bias is to use two independent Monte-Carlo simulations: One simulation is used to estimate the exercise criteria (as a function of some state variable), the other is used to calculate the exercise price based on this exercise criteria. We shall call this the numerical removal of the foresight bias. In this paper we give an exact definition of the foresight bias in closed form and show how to apply an analytical correction for the foresight bias. Our numerical results show that the analytical removal of the foresight bias gives similar results as the standard numerical removal of the foresight bias. The analytical correction allows for a simpler coding and faster pricing, compared to a numerical removal of the foresight bias. Our analysis may also be used as an indication of when to neglect the foresight bias removal altogether. While this is sometimes possible, neglecting foresight bias will break the possibility of parallelization of Monte-Carlo simulation and may be inadequate for Bermudan options with many exercise dates (for which the foresight bias may become a Bermudan option on the Monte-Carlo error) or for portfolios of Bermudan options (for which the foresight bias grows faster than the Monte-Carlo error). In addition to an analytical removal of the foresight bias we derive an analytical correction for the suboptimal exercise due to the uncertainty induced by the Monte-Carlo error. The combined correction for foresight bias (biased high) and suboptimal exercise (biased low) removed the systematic bias even for Monte-Carlo simulations with very small number of paths.

The role of hindsight in foresight: refining strategic reasoning

The purpose of this paper is to deepen understanding of the role that hindsight plays in foresight. The authors argue that the past is not an isolated static state, but one that is intimately connected with the future. However, there are several biases that influence our perceptions and conceptions of the past. These biases act as constraints on our ability to understand the driving forces that emerge from the past, play out through the present and become the critical uncertainties in the future. They could result in misperceptions about events or processes and so may impair foresight methodologies, such as scenario thinking. Such a foresight bias is characterized by a combination of hindsight biases, creeping determinism and searching for information that corresponds to people’s views about both the past and the future. The cognitive linkages between past, present and future are discussed and the role of counter-to-factual analysis is emphasized as an antidote to the foresight bias. Counter-to-factual analysis is both a cognitive process and an analytical reasoning tool applied to the analysis of historical data. Using insights generated from the explorations of counter-to-factual reasoning, the authors present a hindsight-foresight paragon that fortifies current foresight enhancing techniques with counter-to-factual analysis. Two Roads diverged in a yellow wood, And sorry I could not travel both And be one traveler, long I stood To where it bent in the undergrowth… Two roads diverged in a wood, and I– I took the one less traveled by, And that has made all the difference. The Road Not Taken, Robert Frost Men’s curiosity searches past and future And clings to that dimension. But to apprehend The point of intersection of the timeless With time, is an occupation for the saint— T. S. Eliot, from The Dry Salvages