Scribd Logo
Upload

Welcome to Scribd!

  • 47 views

    Teaching Note 96-02: Risk Neutral Pricing in Discrete Time

    Uploaded by

    Veeken Chaglassian
    The document discusses risk neutral pricing in discrete time. It explains that risk neutral pricing does not assume risk neutrality, but rather uses no arbitrage to drive pricing. Under no arbitrage, asset prices can be expressed using "risk neutral probabilities" where the discount rate is the risk-free rate rather than a risky discount factor. This allows pricing without assumptions about investor risk aversion. The document demonstrates how to derive risk neutral probabilities in a simple binomial model and explains their relationship to actual probabilities.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Teaching Note 96-02: Risk Neutral Pricing in Discrete Time

    Uploaded by

    Veeken Chaglassian
    47 views4 pages
    The document discusses risk neutral pricing in discrete time. It explains that risk neutral pricing does not assume risk neutrality, but rather uses no arbitrage to drive pricing. Under no arbitrage, asset prices can be expressed using "risk neutral probabilities" where the discount rate is the risk-free rate rather than a risky discount factor. This allows pricing without assumptions about investor risk aversion. The document demonstrates how to derive risk neutral probabilities in a simple binomial model and explains their relationship to actual probabilities.

    Original Title

    RiskNeutralPricing

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document discusses risk neutral pricing in discrete time. It explains that risk neutral pricing does not assume risk neutrality, but rather uses no arbitrage to drive pricing. Under no arbitrage, asset prices can be expressed using "risk neutral probabilities" where the discount rate is the risk-free rate rather than a risky discount factor. This allows pricing without assumptions about investor risk aversion. The document demonstrates how to derive risk neutral probabilities in a simple binomial model and explains their relationship to actual probabilities.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    47 views4 pages

    Teaching Note 96-02: Risk Neutral Pricing in Discrete Time

    Uploaded by

    Veeken Chaglassian
    The document discusses risk neutral pricing in discrete time. It explains that risk neutral pricing does not assume risk neutrality, but rather uses no arbitrage to drive pricing. Under no arbitrage, asset prices can be expressed using "risk neutral probabilities" where the discount rate is the risk-free rate rather than a risky discount factor. This allows pricing without assumptions about investor risk aversion. The document demonstrates how to derive risk neutral probabilities in a simple binomial model and explains their relationship to actual probabilities.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 4

    TEACHING NOTE 96-02:

    RISK NEUTRAL PRICING IN DISCRETE TIME

    Version date: November 14, 2002 C:\CLASSES\TEACHING NOTES\TN96-02.DOC

    The pricing of derivatives via the risk neutrality argument has been a source of much
    confusion. Those who do not understand the process are led to believe that some mysterious hand-
    waving has been done and that the resulting price is somehow tainted by ignoring the risk. In fact,
    risk neutral pricing makes no such assumption of risk neutrality. It simply uses the assumption of no
    arbitrage rather than risk aversion to drive the pricing process.
    The point is most easily made in the binomial pricing framework. Consider an asset currently
    priced at S whose price one period later can be Su with probability q or Sd with probability 1-q. The
    factors u and d represent one plus the percentage change in the asset price. How would that asset be
    priced in a world of risk averse investors? Under risk aversion any risky asset is priced as the
    discounted value of its future expectation. The expected future price of the asset is qSu + (1- q)Sd.
    Consequently, the current price would be given as

    qSu + (1 - q)Sd
    S=
    1+ k

    where k is the risky discount factor, which will consist of the risk free rate, r, plus a risk premium.
    The rate k is sometimes considered as having arisen out of an asset=s covariance with the market
    portfolio approach (the CAPM) or some alternative model of market equilibrium but it is not
    necessary to assume any such type of equilibrium.
    Given knowledge of S, u and d, however, we can re-state the price of the asset in the
    following manner: Are there values p and 1-p that can be substituted for q and 1-q and permit us to
    change k to the risk-free rate, r. In the absence of arbitrage, the answer is yes.
    Consider the requirements for a world of no arbitrage opportunities and in which there are but
    two assets, the stock and a risk-free bond. If an arbitrage opportunity is possible, then one could
    borrow at the risk-free rate the price of the stock, purchase the stock and guarantee that one period
    later he would earn at least the risk-free rate. This is equivalent to the requirement that u > d > 1 +

    D. M. Chance, TN96-02 1 Risk Neutral Pricing in Discrete Time


    r.1 The factor u is greater than d by definition. The question is whether d is greater than 1+r. If that
    is the case, then such an arbitrage is possible. This is certainly an unreasonable possibility for if it
    existed, everyone would take advantage of it. The interest rate would then have to adjust such that
    the condition u > 1+r > d. Naturally we could also not have 1+r > u for then it would be possible to
    short the stock and use the proceeds to buy the bond, assuring ourselves that at worst our return
    would be 1+r - u. In either case, d > 1+r or 1+r > u, we could guarantee a non-negative payoff
    without investing any of our own money. Consequently, no arbitrage requires that u > 1 + r > d.
    Given that u > 1+r > d, it is possible to find weights p and 1-p that are consistent with the
    statement pu + (1-p)d = 1+r or the equivalent statement, p = (1+r-d)/(u-d). Since this condition
    holds, we are permitted to write the stock price as the following:

    pSu + (1 - p)Sd
    S=
    1+ r

    The proof is simple. Divide the expression above by S and multiply by 1 + r giving 1 + r = pu + (1-
    p)d. This is easily converted into p = (1+r-d)/(u-d).
    We see from the above equation that the price of an asset can be re-stated simply by changing
    the probabilities and discounting at the risk-free rate. The only information we are required to know
    is the volatility (as represented by u and d) and the risk-free rate. We are making no assumptions
    about how investors feel about risk. Indeed what we have done is perfectly compatible with risk
    aversion. If by chance investors were risk averse, the above expression gives precisely the value of
    the asset.
    What we have just seen is called the arbitrage theorem. It simply states that if there are no
    opportunities for arbitrage, then it is possible to state the price of the asset in terms of risk-neutral
    probabilities where the discounting is done at the risk-free rate. The theorem goes two-ways. If it is
    possible to state the price of the asset in the above manner, then there are no arbitrage possibilities.

    1
    The reason we see a A1" in front of the r and not in front of the u and d is that u and d are already specified as 1 plus the up
    and down returns. In some writings, r is already specified as 1 plus the risk-free rate.

    D. M. Chance, TN96-02 2 Risk Neutral Pricing in Discrete Time


    Calling p and 1-p risk neutral probabilities is, however, a source of much confusion.
    Alternatively, these probabilities are sometimes called equivalent martingale probabilities. A
    martingale is a stochastic process in which the expected return of the asset is the current value of the
    asset. Under the equivalent martingale approach, however, the asset price is re-stated so as to be
    already discounted at the risk-free rate. In other words, the S is the expectation of the future price
    already discounted.
    There is a definite relationship between the binomial probability and the actual probability.
    Define the stock=s risk premium as φ = (E(R) - r)/σ where E(R) is the expected return on the stock
    and is defined as E(R) = qu + (1-q)d - 1. The stock=s variance is defined as σ2 = q(1-q)(u-d)2. This
    holds by definition; it is simply the volatility of a one period binomially distributed variable that can
    go up to u or down to d. Substituting these values into the risk premium and noting that p = (1+r-
    d)/(u-d), we obtain the result

    p = q - φ q(1 - q) .

    Thus, the binomial probability is the actual probability minus the risk premium times the square root
    term, which is by definition the volatility of a binomial process. In short, knowing the binomial
    probability, p = (1+r-d)/(u-d), tells us everything we need to know to avoid having to know the risk
    aversion, φ, and the actual probability, q.
    Note that we have not introduced options to the problem. The point is easily established
    without the use of options. When introducing options, however, it is easily shown that a call option
    can be replicated by positions in the asset and the risk-free bond. That being the case, the call is a
    redundant asset. Consequently, introducing a redundant asset does nothing to change the nature of
    the market, provided of course that the redundant asset is properly priced. But that will be true by
    the requirement of no arbitrage.
    Though we have introduced this point in the simple binomial framework, it can also be made,
    albeit with more mathematical complexity in the continuous time models.

    D. M. Chance, TN96-02 3 Risk Neutral Pricing in Discrete Time


    References

    It is difficult to attribute this material to any one source. It has evolved out of a number of prominent
    works such as

    Cox, J. C., S. A. Ross and M. Rubinstein. AOption Pricing: A Simplified Approach.@


    Journal of Financial Economics 7 (September, 1979), 229-263.

    Harrison, M. and D. Kreps. AMartingales and Multiperiod Securities Markets.@


    Journal of Economic Theory, 20 (July, 1979), 381-408.

    Harrison, M. and S. Pliska. AMartingales and Stochastic Integrals in the Theory of


    Continuous Trading.@ Stochastic Processes and Their Applications 11 (1981), 261-
    271.

    Rendleman, R. and B. Bartter. ATwo State Option Pricing.@ The Journal of Finance
    34 (December, 1979), 1092-1110.

    The following is a good treatment, primarily in discrete time:

    Jarrow. R. A. and S. M. Turnbull. Derivative Securities Cincinnati: South-Western


    Publishing (1996), Chs. 5, 6.

    The following is an excellent treatment in continuous time:

    Neftci, S. N. An Introduction to the Mathematics of Financial Derivatives. San


    Diego: Academic Press (1996), Chs. 1, 14, 15.

    D. M. Chance, TN96-02 4 Risk Neutral Pricing in Discrete Time

    You might also like