This paper analyzes the dynamic interaction between selfish and reciprocity firms in the market of homogeneous product. The decisions of both types of firms in respect of their output strategies are investigated under naive, adaptive and generalized expectations. The standard postulate for competitive firms' model has been extended by the assumption that there is a share of reciprocity firms which, unlike selfish firms, maximize both private and social benefits as consumer surplus. It has been proved that the unique Nash equilibrium is stable for all affordable values of parameters in the model with adaptive expectations, and is unstable for the model with naive expectations at sufficiently large number of firms in the market. A special desktop application has been created for animation of model trajectory and demonstration of stable quantity trajectories and bifurcation diagrams of firms' output. Naive expectations of two-thirds of firms result in a state of dynamic chaos in the market leading to degeneration of the existing competition model between the two types of firms. The crucial factor which ensures the stable equilibrium in the market and the ability to predict firms' output is the adaptive approach which takes into account the adaptive expectations of firms planning their product quantity.
In recent years the researchers are renouncing the assumption of perfect rationality as unconditional basis of economic agents’ behavior. The neoclassical ‘rational man’ does not exist in reality; economic agents act according to established rules, without being fully informed and maximizing their own utility [1].
- 503
Karl Polanyi identified the alternative economic organization where social norms are not generated by economic self-interest of the individual. This network of reciprocal relations is based on mutual economic cooperation of efforts and resources between the members of non-economic network, dominated by cultural norms rather than market laws. Under reciprocity relations the exchange donor and recipient can be transposed. So this is a symmetrical relationship of gifts exchange between members of horizontal social networks [2].
This relationship is not regulated by formal institutions but based on informal commitments giving moral right to mutual help and reciprocal exchange on sustainable basis in the long run period. But this is a relationship with minimal risk for participants and the penalty of loss of social capital (reputation and trust) and social isolation. Society supports the stability of the exchange to ensure their survival during crises and wars. Reciprocity is not altruism which does not create reciprocal obligations in quantitative, qualitative or time respects, just vague commitment (e.g., you give, if you can).
Actually, reciprocity relations, commodity exchange and hierarchical subordination exist at the same time. But it is reciprocity that underlies most decentralized corrections of diverse shortcomings and failures of markets and firms. These relations are long-run factors of economic efficiency; they set most social obligations of firms towards individuals without government intervention. No society can exist without reciprocal relationship [3].
Reciprocity or social responsibility implies that the firms not only pursue their selfish goal of increasing profits, but are also ready to sacrifice some of their own profits for the benefit of consumers without direct compensation for it by the state [1]. Such targets can be stipulated by the firms’ desire to get stable profits in the long run rather than maximal short-run profits. Such forward-thinking firms-reciprocators are considered in the model of this paper. Their objective function is a weighted average of the profits and consumer surplus of their market segment.
The real economic processes make a clear demonstration that neoclassical "rational man" is not their subject. In real economy "optimal imperfect decisions" are taken by simple and non-expensive calculations, well adapted to frequent repetitions, to evolution; it is more efficient for perfectly rational firm to perform multiple experiments with quantity to estimate the demand function rather than search for nonrecurrent, instantaneous achieving of equilibrium. New paradigm of nonlinear economics is a mix of qualitative theory of nonlinear dynamical system, optimal control theory, game theory, and theory of stochastic processes [4, 5].
The evolutionary approach and analysis of the dynamics allow to explain why one type of firm ousts another from the market, why sometimes the economic system is stable, but in other cases is unstable [6, 7]. If the system has multiple equilibria, the dynamics and evolution is the selection mechanism of best equilibrium according to certain criteria [8]. The evolutionary process is analogous to social learning. An example of its application is the pricing mechanisms for auctions that occur in agents social networks, e-commerce and trade through the Internet [9, 10].
The study of the evolution of the markets with the strategic interaction usually uses the following assumptions: (a) two firms or two types of similar firms operate in the market; (b) firms produce homogeneous goods in quantities of x1(t ) and x2 (t ) ; (c) no firm knows the rivals’ quantities; (d) the firms seek to predict the output of the competitors using the adaptive scheme.
Planning of quantity for the following period firms resolve optimization problem: MaxПi (xi ; xej (t 1)) , where Пi is the objective function of firm i, xej (t 1) - expected quantity of a competitor j ( i, j 1, 2 ).
Examples of bounded rationality of firms are: ignoring the impact of competitors' actions on their own output (local monopoly approximation LMA), naive expectation (assumption of unchangeable behavior of competitors for a long time and using x j (t) instead xej (t 1) ). [8] Of course both decision making approaches (adaptive and naive, bounded rational) coexist in the market with a certain probability.
Analysis of nonlinear oligopoly with heterogeneous players reveals that a higher degree of product differentiation may destabilize the Cournot-Nash equilibrium. Authors showed that a cascade of flip bifurcation may lead to periodic cycles and chaotic motions [11]. Stability conditions of Nash equilibrium and complex dynamics are also studied for heterogeneous duopoly with isoelastic demand function. For such heterogeneous players a cascade of flip bifurcation leads to periodic cycles and chaos and the Neimark-Sacker bifurcation generates attractive invariant closed curve [12].
Such scheme serves as the basis for mathematical model of this paper, which distinguishes from the other models in that: (a) firms use more than one way of decision-making, and combine different ones; (b) except their own selfish interests, firms take into account social ones.
The paper goal is to consider the impact of naive, adaptive and generalized expectations of egoist and reciprocator firms on stability of equilibrium and the conditions of transition to dynamic chaos in the numerical experiment using a specially designed desktop application.
The paper is organized as follows: part 2 describes two-dimensional market model with naïve and adaptive expectations; part 3 is devoted to dynamics model in general case; part 4 demonstrates C#-application model for numerical investigation; part 5 concludes. 2
We consider the market of homogeneous product, which consists of n firms, including k identical firms-reciprocators, each of them producing x units of product and n – k identical selfish firms, each of them producing y units of product. Thus, the industry quantity of the two types of firms is Q k x (n k ) y . Product price P in the market is given by the inverse market demand function P P(Q) ( b 0 ).
The objective function of a firm-egoist is profit Y (P v) y , where v is the firm’s costs per unit in the market. Firm-reciprocator maximizes both its own profit X (P v) x and consumer surplus CS of its own market segment: b
Q Q CS k P(q)dq PQ , where is the parameter defining the segment of the market, which the reciprocator firm believes its own and optimizes it ( 0 k ), is the minimal technologically possible quantity of product. Then
CS k b ln Q Qb Q bk ln Q 1 bk ln Q , where e . The specific choice of ε does not affect the dynamics of the model because the objective function, as any potential, is set up to an arbitrary constant accuracy, so further we will write ε instead of ˆ . Then the objective function of firmreciprocator is: Consider the dynamics of this two-dimensional model in discrete time t = 0, 1,...; where xt, yt are the outputs at time t of reciprocator and egoist firm, respectively. On the basis of these values at time t each firm finds the optimal value for its own production quantity in the next moment t 1 , maximizing its objective function. It distinguishes this model, in which the firm responds to changes in output of both their and other types of firms from traditional competition models, where one type of firm responds to changes in other types only. So each selfish firm is looking for such value of yt1 at which it maximizes its own profits, suggesting that SR firms and the other n k 1 PI firms leave their quantities unchanged:
b
Y v yt1 . (
yt1 k xt (n k 1) yt
Obviously, the maximum point for yt1 is found from the condition
0 , whence:
v( yt1 kxt (n k 1) yt )2 b(kxt (n k 1) yt ) .
(
yt1 bv (kxt (n k 1) yt kxt (n k 1) yt . (
. Then v(xt1 (k 1)xt (n k ) yt )2 b((k 1)xt (n k ) yt ) 1 bk (xt1 (k 1)xt (n k ) yt ) .
Assuming z xt1 (k 1)xt (n k) yt , we present (
v vk
Hence, in view of (
Equilibrium Conditions for the Model with Naive Expectations
In the Nash equilibrium point xt+1=xt=x, yt+1 =yt=y for all t = 0, 1, … . Therefore, at
this point, by (
From the last equation we obtain x y 1 n k ) y , i.e. the response functions of both types of firms are k 2 1
x (1 respectively: x k (1 ) (n k ) y , y (2 1) k x
(2 1) k k (1 ) (n k)
To calculate the coordinates of a fixed point, we substitute the expression of y
through x in the first equation (
k
b 2 1 1 n k
x* 1 1 ,
vn n k
(
b 2 1 2 1 y* vn 1 n .
However, is this point stable?
Proposition 2. For any set b , v 0 and ( 0 1 ) Nash equilibrium (
The destabilizing role of number of players n is well known for the evolution of
firms’ strategies in oligopoly games [8]. However, in this case, according to
calculations, point (
Proof. We show that in dynamic system (
((k 1)x * (n k 1) y*) b v
Jxy J yy k 3 at a certain 0 , which guarantees that the factors n 4 b v b v 1 and 2 b ((k 1)xt (n k) yt d 2 2 b
v v
zero for all possible n, k, b, v> 0 and α (0≤α≤1), Q.E.D.
((k 1)xt (n k 1) yt
1 do not equal
(
Dynamic Model Equations with Adaptive Expectations Since all selfish firms are assumed as identical, it is natural to suggest that they have the same planning at moment t , so their production quantities yt1 at moment t 1 will be equal too. Given these expectations, each selfish firm is looking for such value yt1 at which it obtains the highest profit, suggesting that production quantity of SR firms will remain unchanged: Obviously, the maximum point for yt1 is found from the condition Y 0 , which yt1 gives us:
b Y kxt (n k ) yt1
v yt1 .
kxt (n k) yt1 2
kxt .
b
v
b
v
Then kxt (n k ) yt1
kxt , from here response function of PI firms is:
(n k) yt1 b kxt kxt . (
v
Similarly, firm-reciprocator naturally expects that the quantity of production of all these firms at moment t 1 would be the same. Based on this expectation, each firmreciprocator finds the value of xt1 at which the objective function is maximal, assuming that the output of PI firms does not change: X b xt1 vxt1 (1 ) b kxt1 (n k ) yt k
kxt1 (n k ) yt . ln
In the Nash equilibrium point xt+1=xt=x, yt+1=yt=y for all t 0,1,... . Therefore, at this
point in view of (
(18a) As before, without loss of generality, let 1 , otherwise we can override the share of profit as
J xy xt , obviously, J xx J yy 0 where trJ J xx J yy 0 . Thus, to test Shur conditions it is sufficient to establish that det J 1 . But at point (18) y* and therefore b v(n k) 2 ((1 ) )2 (2 )2 kJxy Consequently, det J J xx J yy J xy J yx 0 , Q.E.D. function P P(Q)
The price of product P in the market is given by the inverse market demand b
( b 0 ), and the price is not less than a cent, i.e. P 0.01 .
Q Therefore, the product quantity of each firm-reciprocator is x . Similarly, the Suppose that in planning under the given market model adaptive expectations are used with probability p , naïve ones - with probability q 1 p . Then the profit function for a typical (representative) firm-egoist has the form:
b Y yt1 kxt p(n k 1) yt1 q(n k 1) yt v yt1 , and the objective function for the representative firm-reciprocator
b X xt1 p(k 1)xt1 q(k 1)xt (n k) yt xt1 vxt1 , (1 ) b ln( xt1 p(k 1)xt1 q(k 1)xt (n k) yt ) .
k
Obviously, for p 0 ( q 1 ) objective functions Y and X are consistent with the
results of naive model (
(kxt q(n k 1) yt yt1 (1 p(n k 1)) b (kxt q(n k 1) yt k xt1 (n k 1) q yt (22) v The maximum point xt1 for the objective function X is found from the first order
X 0 . Forth without loss of generality we assume here 1 , xt1 otherwise as above we redefine the share of profit as
z2X bv ((n k ) yt (k 1) qxt ) 1 b(1 pvk(k 1)) zX (23) Thus, in view of (22), we obtain the dynamics model of equations system of in the general case:
y p n k q(1 1 n k k ) x (1 p(k 1)) 21 , where response function in this case p(n k) q( (1 ) n k )
k x G . (26) y (2 1)(1 p(k 1)) To calculate the coordinates of the fixed point we substitute from (26) expression for x in the first equation of (25) y2 (kG (n k))2 b y(kG q(n k 1)) . Hence y v Proposition 5. There is unique Nash equilibrium point in a general dynamical system (24):
b (kG q(n k 1)) y* v
x* Gy* v (kG (n k))2 (k (1/ G)(n k ))2 where the function G G( p, q, n, k, ) is given in (26).
Proposition 6. For p 0 ( q 1 ) the equilibrium point (x*; y*) coincides with point
(
(27) b (k q(1/ G)(n k 1))
C# - Application Model for Numerical Investigation C# window application Model has been created specifically for the numerical investigation of the model of this paper, using a graphical interface of C# system libraries System.Drawing and System.Windows.Forms. Note that all the calculations associated with the model, are localized in the method calc of the application Model that makes it easy to modify the equations of the model and use the Model to study the other two-dimensional dynamical systems. Fig. 1 shows the application window.
The right side presents 6 kinds of graphs displayed by the application; their examples are set forth in the paper. Selected switch indicates that here the graph of trajectory x(t) is selected. On the left side counters allow us to specify the parameters of the model and the initial values of the trajectory. After their setting the calculation results of the iterations’ coordinates below and their image in the center of the window. This displays an animation of a selected path, the number of iterations been set on the scroll bar above. Pressing the button Model view left displays information about the model, its equations and parameter information.
With the increasing probability of naive expectations q, that is with decreasing p,
the market becomes unstable, evolving from simple dynamics (15) with a single
stable equilibrium point to the unpredictable behavior of system (
As we can see from the graph, the path quickly converges to the equilibrium value y*≈2.488. The graph for the trajectory of firm-reciprocator xm on y-axis is similar. The equilibrium value of x* is about 6.72. Let us consider the graph of the trajectory for the same parameters except q. Now q = 0.55 (Fig. 3). It still has stable Nash equilibrium, but 100 iterations does not suffice for convergence. Further, let q = 0.6 (Fig. 4).
Fig. 4. Bifurcation quantity trajectory of selfish firm under probability of naive expectations q 0.6 As we can see, bifurcation occurred, and instead of equilibrium point there was a steady cycle, where values of ym are approaching the point of y*≈4 for even m and the point of y*≈1 for odd m. By doubling the lag between iterations only even or only odd iterations will be considered, and thus either point y*≈4, or y*≈1 respectively would be the equilibrium steady state.
Stable cycle has four cycles for q=0.64 (fig. 5). There was a new cycle doubling bifurcation. Calculations show that with increasing parameter q doubling bifurcation cycle continues, following Sharkovskii’s scale. According to this scale, when q≈0.675 there is the state of dynamic chaos (fig. 6). Similarly, the graph of product xm on yaxis by firm-reciprocator looks like trajectory of a selfish firm.
Fig. 5. Doubling bifurcation cycle of quantity by selfish firm under probability of naive expectations q 0.64 Fig. 6. The state of dynamic chaos of quantity by selfish firms under probability of naive expectations q 0.675
Note that the ratio between the quantity of output by selfish firms and reciprocators remains almost unchanged. It is demonstrated in the graph of fig. 9, where each iteration on x-axis shows the value of output by firms-reciprocators xm, and the vertical axis - the appropriate output of quantity ym of selfish firms (fig. 7.). In detail the process of loss of stability and transition to chaos of dynamic system (24) can be presented in the following bifurcation diagram (fig. 8). probability of naive expectations ( q ) in a general dynamical system
Here the horizontal axis represents the parameter value of q multiplied by 10. The ordinate values quantity volumes of selfish firm on stable cycle, multiplied by 0.3. This rescaling is done for the sake of clarity. The values of the other parameters are the same as above. The bifurcation diagram, where on vertical axis are placed the values of output of firms-reciprocators xm looks similar.
As noted in numerical simulations, the bifurcation may be interpreted as separation of equilibrium into several ways, one of which is selected by the market due to evolution of firms’ strategies, such as repeated interactions and adaptations. Numerical experiments with n firms as the variable parameter are analogous to those described above. 5
Thus, we have designed the strategic model of cooperation between the two types of firms in the market of homogeneous product, where reciprocator and selfish firms plan their output using the adaptive approach with probability p and naïve (bounded rationality) one with a probability of q = 1-p, which distinguishes this model from existing analogues, where each type of firm adheres to one strategy rather than their combination and maximizes only its own profit rather than social welfare.
Desktop C# application Model using a graphical interface to animate the model trajectories has been created specifically for the numerical investigation of the model.
It has been proved that in the model with adaptive expectations the unique Nash equilibrium in a dynamic system is stable for all possible values of the parameters. The trajectories of the dynamical system converge to the fixed point for any possible initial values. In the model with naive expectations the unique Nash equilibrium is unstable for sufficiently large values of n for all possible values of other parameters. According to the calculations, this point is unstable even at n ≥ 5.
As a result of numerical experiment we have found that bifurcations of cycle doubling occur with an increase in naive expectations. This bifurcation can be interpreted as separation of equilibrium state into several ways, one of which is selected by the market in the evolution of firms’ strategies. If two-thirds of firms use naive expectation (q≈0.675), then in accordance with the Sharkovskii scale there appears the state of dynamic chaos in the market, leading to degeneration of the existing competition model between two types of firms.
Thus, the crucial factor, which ensures sustainable equilibrium in the market and the ability to predict the product quantity of firms, is the adaptive approach, i.e. the one taking into account adaptive expectations of the firms when they plan their production.
Similar results are obtained if instead of q we use parameter n - number of firms in the market, where system also moves from stability to chaos if n increases.